Showing papers on "Fault (geology) published in 2021"

WaveletKernelNet: An Interpretable Deep Neural Network for Industrial Intelligent Diagnosis

Abstract: Convolutional neural network (CNN), with the ability of feature learning and nonlinear mapping, has demonstrated its effectiveness in prognostics and health management (PHM). However, an explanation on the physical meaning of a CNN architecture has rarely been studied. In this article, a novel wavelet-driven deep neural network, termed as WaveletKernelNet (WKN), is presented, where a continuous wavelet convolutional (CWConv) layer is designed to replace the first convolutional layer of the standard CNN. This enables the first CWConv layer to discover more meaningful kernels. Furthermore, only the scale parameter and translation parameter are directly learned from raw data at this CWConv layer. This provides a very effective way to obtain a customized kernel bank, specifically tuned for extracting defect-related impact component embedded in the vibration signal. In addition, three experimental studies using data from laboratory environment are carried out to verify the effectiveness of the proposed method for mechanical fault diagnosis. The experimental results show that the accuracy of the WKNs is higher than CNN by more than 10%, which indicate the importance of the designed CWConv layer. Besides, through theoretical analysis and feature map visualization, it is found that the WKNs are interpretable, have fewer parameters, and have the ability to converge faster within the same training epochs.

Modified Stacked Auto-encoder Using Adaptive Morlet Wavelet for Intelligent Fault Diagnosis of Rotating Machinery

Abstract: Intelligent fault diagnosis techniques play an important role in improving the abilities of automated monitoring, inference, and decision-making for the repair and maintenance of machinery and processes In this paper, a modified stacked auto-encoder (MSAE) that uses adaptive Morlet wavelet is proposed to automatically diagnose various fault types and severities of rotating machinery Firstly, the Morlet wavelet activation function is utilized to construct an MSAE to establish an accurate nonlinear mapping between the raw nonstationary vibration data and different fault states Then, the nonnegative constraint is applied to enhance the cost function to improve sparsity performance and reconstruction quality Finally, the fruit fly optimization algorithm (FOA) is used to determine the adjustable parameters of the Morlet wavelet to flexibly match the characteristics of the analyzed data The proposed method is used to analyze the raw vibration data collected from a sun gear unit and a roller bearing unit Experimental results show that the proposed method is superior to other state-of-the-art methods

The generation of large earthquakes

Abstract: Despite decades of observational, laboratory and theoretical studies, the processes leading to large earthquake generation remain enigmatic. However, recent observations provide new promising perspectives that advance knowledge. Here, we review data on the initiation processes of large earthquakes and show that they are multiscale and diverse, involving localization of deformation, fault heterogeneities and variable local loading rate effects. Analyses of seismic and geodetic data reveal evidence for regional weakening by earthquake-induced rock damage and progressive localization of deformation around the eventual rupture zones a few years before some large earthquakes. The final phase of deformation localization includes, depending on conditions, a mixture of slow slip transients and foreshocks at multiple spatial and temporal scales. The evolution of slip on large, localized faults shows a step-like increase that might reflect stress loading by previous failures, which can produce larger dynamic slip, in contrast to the smooth acceleration expected for a growing aseismic nucleation phase. We propose an integrated model to explain the diversity of large earthquake generation from progressive volumetric deformation to localized slip, which motivates future near-fault seismic and geodetic studies with dense sensor networks and improved analysis techniques that can resolve multiscale processes. The processes leading to large earthquakes remain enigmatic. Using detailed seismic and geodetic data, this Review examines how tectonic deformation and evolving fault behaviour initiate large earthquakes, and proposes an integrated model accounting for the diversity of observations.

Sealing mechanisms in volcanic faulted reservoirs in Xujiaweizi extension, northern Songliao Basin, northeastern China

Abstract: Distribution of volcanic reservoirs in Xujiaweizi half graben is controlled by both faults and sealing layers where their integrity could have been compromised. This paper documents the distribution of sealing units of the Denglouku and Yingcheng Formations based on seismic and well-log data to delineate all of the reservoirs with proper cap rocks within the volcanic Yincheng Formation via the relationship between the fault system and reservoir rocks. Based on the regional sealing effect of Denglouku Formation, two hydrocarbon traps are identified: the lower primary and upper secondary gas reservoirs where the formation with a mudstone percentage of more than 50% within the second member seals the primary reservoirs. The upper secondary trap is controlled by faults that have been activated during the structural reversal phase, causing the regional sealing layers in the Denglouku Formation to get displaced with a cap rock juxtaposition thickness of less than 35 m. This has created a series of fault-seal, dual-control reservoirs. The local seals within the Yingcheng Formation consist of mudstone, tight volcanic rock, and clayey breccia covering each volcanic eruption cycle. These local seals separate the volcanic gas reservoirs with a minimum thickness of 20 m. The local top seals in this tectonically active zone were placed on the hanging wall via the juxtaposition of the reservoir and overlying mudstone and/or clayey breccia. It was concluded that gas has migrated vertically through the faults and accumulated in the fault-controlled traps where sharp changes in the lithology (juxtaposition) form the seal, whereas the gas-water contact is controlled by the depth of the reservoir rock. Finally, this study concludes that the primary reservoirs are distributed in a sinusoidal configuration around the fault zone because of the dolphin effect of the Xuzhong strike-slip fault system that has connected the source and reservoir rocks.

A new dynamic model and transfer learning based intelligent fault diagnosis framework for rolling element bearings race faults: Solving the small sample problem.

Abstract: Intelligent fault diagnosis of rolling element bearings gains increasing attention in recent years due to the promising development of artificial intelligent technology. Many intelligent diagnosis methods work well requiring massive historical data of the diagnosed object. However, it is hard to get sufficient fault data in advance in real diagnosis scenario and the diagnosis model constructed on such small dataset suffers from serious overfitting and losing the ability of generalization, which is described as small sample problem in this paper. Focus on the small sample problem, this paper proposes a new intelligent fault diagnosis framework based on dynamic model and transfer learning for rolling element bearings race faults. In the proposed framework, dynamic model of bearing is utilized to generate massive and various simulation data, then the diagnosis knowledge learned from simulation data is leveraged to real scenario based on convolutional neural network (CNN) and parameter transfer strategies. The effectiveness of the proposed method is verified and discussed based on three fault diagnosis cases in detail. The results show that based on the simulation data and parameter transfer strategies in CNN, the proposed method can learn more transferable features and reduce the feature distribution discrepancy, contributing to enhancing the fault identification performance significantly.

Fault Diagnosis of Hydraulic Systems Based on Deep Learning Model With Multirate Data Samples.

Abstract: Hydraulic systems are a class of typical complex nonlinear systems, which have been widely used in manufacturing, metallurgy, energy, and other industries. Nowadays, the intelligent fault diagnosis problem of hydraulic systems has received increasing attention for it can increase operational safety and reliability, reduce maintenance cost, and improve productivity. However, because of the high nonlinear and strong fault concealment, the fault diagnosis of hydraulic systems is still a challenging task. Besides, the data samples collected from the hydraulic system are always in different sampling rates, and the coupling relationship between the components brings difficulties to accurate data acquisition. To solve the above issues, a deep learning model with multirate data samples is proposed in this article, which can extract features from the multirate sampling data automatically without expertise, thus it is more suitable in the industrial situation. Experiment results demonstrate that the proposed method achieves high diagnostic and fault pattern recognition accuracy even when the imbalance degree of sample data is as large as 1:100. Moreover, the proposed method can increase about 10% diagnosis accuracy when compared with some state-of-the-art methods.

Intelligent Mechanical Fault Diagnosis Using Multi-Sensor Fusion and Convolution Neural Network

Abstract: Diagnosis of mechanical faults in manufacturing systems is critical for ensuring safety and saving costs. With the development of data transmission and sensor technologies, measuring systems can acquire massive amounts of multi-sensor data. Although Deep-Learning (DL) provides an end-to-end way to address the drawbacks of traditional methods, it is necessary to do deep research on an intelligent fault diagnosis method based on Multi-Sensor Data. In this project, a novel intelligent diagnosis method based on Multi-Sensor Fusion (MSF) and Convolutional Neural Network (CNN) is explored. Firstly, a Multi-Signals-to-RGB-Image conversion method based on Principal Component Analysis (PCA) is applied to fuse multi-signal data into three-channel RGB images. Then, an improved CNN with residual networks is proposed, which can balance the relationship between computational cost and accuracy. Two datasets are used to verify the effectiveness of the proposed method. The results show the proposed method outperforms other DL-based methods in terms of accuracy.

Aftershock sequence relocation of the 2021 MS7.4 Maduo Earthquake, Qinghai, China

Abstract: The 2021 Qinghai Maduo MS7.4 earthquake was one of the strongest earthquakes that occurred in the Bayan Har block of the Tibetan Plateau during the past 30 years, which spatially filled in the gap of strong earthquake in the eastern section of the northern block boundary. In this study, the aftershock sequence within 8 days after the mainshock was relocated by double difference algorithm. The results show that the total length of the aftershock zone is approximately 170 km; the mainshock epicenter is located in the center of the aftershock zone, indicating a bilateral rupture. The aftershocks are mainly distributed along NWW direction with an overall strike of 285°. The focal depth profiles indicate that the seismogenic fault is nearly vertical and dips to southwest or northeast in different sections, indicating a complex geometry. There is an aftershock gap located to the southeast of the mainshock epicenter with a scale of approximately 20 km. At the eastern end of the aftershock zone, horsetail-like branch faults show the terminal effect of a large strike-slip fault. There is a NW-trending aftershock zone on the north side of the western section, which may be a branch fault triggered by the mainshock. The location of the aftershock sequence is close to the eastern section of the Kunlun Mountain Pass-Jiangcuo (KMPJ) fault. The sequence overlaps well with surface trace of the KMPJ fault. We speculate that the KMPJ fault is the main seismogenic fault of the MS7.4 Maduo earthquake.

A Single-Side Neural Network-Aided Canonical Correlation Analysis With Applications to Fault Diagnosis

Abstract: Recently, canonical correlation analysis (CCA) has been explored to address the fault detection (FD) problem for industrial systems. However, most of the CCA-based FD methods assume both Gaussianity of measurement signals and linear relationships among variables. These assumptions may be improper in some practical scenarios so that direct applications of these CCA-based FD strategies are arguably not optimal. With the aid of neural networks, this work proposes a new nonlinear counterpart called a single-side CCA (SsCCA) to enhance FD performance. The contributions of this work are four-fold: 1) an objective function for the nonlinear CCA is first reformulated, based on which a generalized solution is presented; 2) for the practical implementation, a particular solution of SsCCA is developed; 3) an SsCCA-based FD algorithm is designed for nonlinear systems, whose optimal FD ability is illustrated via theoretical analysis; and 4) based on the difference in FD results between two test statistics, fault diagnosis can be directly achieved. The studies on a nonlinear three-tank system are carried out to verify the effectiveness of the proposed SsCCA method.

Subdomain Adaptation Transfer Learning Network for Fault Diagnosis of Roller Bearings

Abstract: Due to the data distribution discrepancy, fault diagnosis models, trained with labeled data in one scene, likely fails in classifying by unlabeled data acquired from the other scenes. Transfer learning is capable to generalize successful application trained in one scene to the fault diagnosis in the other scenes. However, the existing transfer methods do not pay much attention to reduce adaptively marginal and conditional distribution biases, and also ignore the degree of contribution between both biases and among network layers, which limit classification performance and generalization in reality. To overcome these weaknesses, we established a new fault diagnosis model, called subdomain adaptation transfer learning network (SATLN). Firstly, two convolutional building blocks were stacked to extract transferable features from raw data. Then, the pseudo label learning was amended to construct target subdomain of each class. Furthermore, a sub-domain adaptation was combined with domain adaptation to reduce both marginal and conditional distribution biases simultaneously. Finally, a dynamic weight term was applied for adaptive adjustment of the contributions from both discrepancies and each network layers. The SATLN method was tested with six transfer tasks. The results demonstrate the effectiveness and superiority of the SATLN in the cross-domain fault diagnosis field.

Semi-supervised meta-learning networks with squeeze-and-excitation attention for few-shot fault diagnosis.

Abstract: In the engineering practice, lacking of data especially labeled data typically hinders the wide application of deep learning in mechanical fault diagnosis. However, collecting and labeling data is often expensive and time-consuming. To address this problem, a kind of semi-supervised meta-learning networks (SSMN) with squeeze-and-excitation attention is proposed for few-shot fault diagnosis in this paper. SSMN consists of a parameterized encoder, a non-parameterized prototype refinement process and a distance function. Based on attention mechanism, the encoder is able to extract distinct features to generate prototypes and enhance the identification accuracy. With semi-supervised few-shot learning, SSMN utilizes unlabeled data to refine original prototypes for better fault recognition. A combinatorial learning optimizer is designed to optimize SSMN efficiently. The effectiveness of the proposed method is demonstrated through three bearing vibration datasets and the results indicate the outstanding adaptability in different situations. Comparison with other approaches is also made under the same setup and the experimental results prove the superiority of the proposed method for few-shot fault diagnosis.

Intelligent Fault Diagnosis for Large-Scale Rotating Machines Using Binarized Deep Neural Networks and Random Forests

Abstract: Recently, deep neural network (DNN) models work incredibly well, and edge computing has achieved great success in real-world scenarios, such as fault diagnosis for large-scale rotational machinery. However, DNN training takes a long time due to its complex calculation, which makes it difficult to optimize and retrain models. To address such an issue, this work proposes a novel fault diagnosis model by combining binarized DNNs (BDNNs) with improved random forests (RFs). First, a BDNN-based feature extraction method with binary weights and activations in a training process is designed to reduce the model runtime without losing the accuracy of feature extraction. Its generated features are used to train an RF-based fault classifier to relieve the information loss caused by binarization. Second, considering the possible classification accuracy reduction resulting from those very similar binarized features of two instances with different classes, we replace a Gini index with ReliefF as the attribute evaluation measure in training RFs to further enhance the separability of fault features extracted by BDNN and accordingly improve the fault identification accuracy. Third, an edge computing-based fault diagnosis mode is proposed to increase diagnostic efficiency, where our diagnosis model is deployed distributedly on a number of edge nodes close to the end rotational machines in distinct locations. Extensive experiments are conducted to validate the proposed method on the data sets from rolling element bearings, and the results demonstrate that, in almost all cases, its diagnostic accuracy is competitive to the state-of-the-art DNNs and even higher due to a form of regularization in some cases. Benefited from the relatively lower computing and storage requirements of BDNNs, it is easy to be deployed on edge nodes to realize real-time fault diagnosis concurrently.

A multi-source dense adaptation adversarial network for fault diagnosis of machinery

Abstract: Deep learning (DL) theory has made great progress in the field of intelligent fault diagnosis, and the development of domain adaptation has greatly promoted fault diagnosis under polytropic working conditions (PWC). Extensive studies have been conducted to solve the problem of fault diagnosis under PWC. However, the existing fault diagnosis methods based on domain adaptation have the following shortcomings. First, multi-source information fusion is rarely considered. Second, the utilization of inherent labels is also insufficient in classification problems. To deal with the above problem, a novel multi-source dense adaptation adversarial network (MDAAN) is proposed, which leverages multi-sensor vibration information and classification label information. Specifically, the frequency spectrum of multi-sensor data are firstly employed to make full use of fault information. Afterwards, the dense convolution and fusion convolution blocks are used for deep feature extraction and fusion. Finally, a joint loss function is reconstructed under the framework of unsupervised learning, which considers the distribution differences of the features and the label information simultaneously. The experimental results from various working conditions, including still distant working conditions, all demonstrate that the proposed method can achieve state-of-the-art performances, which has shown great promise as an intelligent fault diagnosis method.

Large-Scale Fracture Systems Are Permeable Pathways for Fault Activation During Hydraulic Fracturing

Abstract: The association of induced seismicity with hydraulic-fracturing (HF) operations for shale gas extraction is well-established (e.g., Atkinson et al., 2016; Bao & Eaton, 2016; Clarke et al., 2019; Verdon & Bommer, 2020; Zoback & Kohli, 2019). The potential socio-economic impact of hydraulic fracturing-induced seismicity worldwide can be high (Atkinson et al., 2020), as exemplified by a ML 5.7 event in China in 2018, which resulted in injuries and millions of dollars in damages (Lei et al., 2019). Kao et al. (2018) identified at least five instances in western Canada of M > 4.0 induced events, while other notable cases of hydraulic fracturing-induced seismicity have been documented in Ohio (Friberg et al., 2014; Skoumal et al., 2015), Oklahoma (Holland, 2013), and the UK (Clarke et al., 2019; Kettlety et al., 2020a). For many published case studies of hydraulic fracturing-induced seismicity, events were recorded using regional seismograph networks at distances of 10s of km (or more), or local monitoring was installed after-the-fact once seismicity had started (e.g., Clarke et al., 2014; Darold et al., 2014; Friberg et al., 2014; Schultz et al., 2015a, 2015b; Skoumal et al., 2015; Wang et al., 2016). With such limitations, further investigation into the causative mechanisms of induced seismicity is often hindered, meaning that competing hypotheses cannot always be conclusively tested (e.g., Deng, Liu, & Harrington, 2016; Schultz et al., 2017). Abstract Induced seismicity due to fluid injection, including hydraulic fracturing, is an increasingly common phenomenon worldwide; yet, the mechanisms by which hydraulic fracturing causes fault activation remain unclear. Here we show that preexisting fracture networks are instrumental in transferring fluid pressures to larger faults on which dynamic rupture occurs. Studies of hydraulic fracturing-induced seismicity in North America have often used observations from regional seismograph networks at distances of 10s of km, and as such lack the resolution to answer some of the key questions about triggering mechanisms. To carry out a more detailed analysis of the mechanisms of fault activation, we use data from a dense sensor array located at a hydraulic-fracturing site in Alberta, Canada. The spatiotemporal distribution of event hypocenters, coupled with measurements of seismic anisotropy, reveal the presence of preexisting fracture corridors that allowed communication of fluid-pressure perturbations to larger faults, over distances of 1 km or more. The presence of preexisting permeable fracture networks can significantly increase the volume of rock affected by the pore-pressure increase, thereby increasing the probability of induced seismicity. This study demonstrates the importance of understanding the connectivity of preexisting natural fractures for assessing potential seismic hazards associated with hydraulic fracturing of shale formations and offers a detailed case exposition of induced seismicity due to hydraulic fracturing.

Transfer Relation Network for Fault Diagnosis of Rotating Machinery With Small Data.

Abstract: Many deep-learning methods have been developed for fault diagnosis. However, due to the difficulty of collecting and labeling machine fault data, the datasets in some practical applications are relatively much smaller than the other big data benchmarks. In addition, the fault data come from different machines. Therefore, on some occasions, fault diagnosis is a multidomain problem with small data, where satisfactory transfer performance is difficult to obtain and has been rarely explored from the few-shot learning viewpoint. Different from the existing deep transfer learning solutions, a novel transfer relation network (TRN), combining a few-shot learning mechanism and transfer learning, is developed in this study. Specifically, the fault diagnosis problem has been treated as a similarity metric-learning problem instead of solely feature weighted classification. A feature net and a relation net have been, respectively, constructed for feature extraction and relation computation. The Siamese structure has been borrowed to extract the features of the source and the target domain samples with shared weights. Multikernel maximum mean discrepancy (MK-MMD) is employed on several higher layers with different tradeoff parameters to enable an efficient domain feature transfer considering different feature properties. To implement efficient diagnosis based on small data, an episode-based few-shot training strategy is adopted to train TRN. Average pooling has been adopted to suppress the noise influence from the vibration sequence which turns out to be important for the success of time sequence-based fault diagnosis. Transfer experiments on four datasets have verified the superior performance of TRN. A significant improvement of classification accuracy has been made compared with the state-of-the-art methods on the adopted datasets.

Slip rate deficit and earthquake potential on shallow megathrusts

Abstract: Most destructive tsunamis are caused by seismic slip on the shallow part of offshore megathrusts. The likelihood of this behaviour is partly determined by the interseismic slip rate deficit, which is often assumed to be low based on frictional studies of shallow fault material. Here, we present a new method for inferring the slip rate deficit from geodetic data that accounts for the stress shadow cast by frictionally locked patches, and show that this approach greatly improves our offshore resolution. We apply this technique to the Cascadia and Japan Trench megathrusts and find that, wherever locked patches are present, the shallow fault generally has a slip rate deficit between 80 and 100% of the plate convergence rate, irrespective of its frictional properties. This finding rules out areas of low kinematic coupling at the trench considered by previous studies. If these areas of the shallow fault can slip seismically, the global tsunami hazard could be higher than currently recognized. Our method identifies critical locations where seafloor observations could yield information about frictional properties of these faults so as to better understand their slip behaviour. Shallow parts of megathrusts up-dip of locked patches generally have a high slip rate deficit, which could mean tsunami hazard has been underestimated, according to a stress-constrained inversion of geodetic data.

Airgap Search Coil based Identification of PM Synchronous Motor Defects

Abstract: On-line detection of rotor and load faults in permanent magnet synchronous motors (PMSM) based on spectrum analysis of current or vibration is not capable of identifying the root cause of the fault, since all faults produce identical fault signatures. The sensitivity of fault detection also depends on the motor and controller design making fault detection unreliable. In addition, operation under variable frequency and load limits the effectiveness of spectrum analysis based methods. In this paper, airgap flux monitoring is investigated as an alternative for providing reliable identification of rotor and load faults in PMSMs. Based on the analysis of airgap flux under partial and uniform demagnetization, dynamic eccentricity and load unbalance, an airgap search coil voltage based method for detection and classification of the faults is proposed. The claims made in the paper are verified through experimental testing on an IPMSM under emulated fault conditions along with a comparison to vibration and current spectra-based detection. It is shown that the proposed method provides reliable on-line identification of the faults for cases where conventional spectrum analysis based methods fail.

Technical Review of Dual Inverter Topologies for More Electric Aircraft Applications

Abstract: Electric drives are an essential part of more electric aircraft (MEA) applications with ever-growing demands for high power density, high performance, and high fault-tolerant capability. High-speed motor drives can fulfil those needs, but their speeds are subject to the relatively low DC-link voltage adopted by MEA. The power inverters are thus expected to efficiently and effectively manage that limited voltage. A recently popular topology is represented by the dual inverters. They are featured by inherited fault tolerance, a high DC-link voltage utilization and an excellent output power profile. This paper aims to present a comprehensive review of different structures based on the dual inverter. To meet the stringent requirements of MEA applications, three performance aspects, including the voltage utilization, the inverter output quality and the fault-tolerant capability, are selected. Based on the chosen performance metrics, the key features of adopting dual inverter topologies against other converter selections are explicitly demonstrated. Finally, a practical guideline for choosing suitable dual inverters for different MEA applications is provided.

Bookshelf Kinematics and the Effect of Dilatation on Fault Zone Inelastic Deformation: Examples From Optical Image Correlation Measurements of the 2019 Ridgecrest Earthquake Sequence

Abstract: The 2019 Ridgecrest earthquake sequence initiated on July 4th with a series of foreshocks, including a Mw 6.4 event, that culminated a day later with the Mw 7.1 mainshock and resulted in rupture of a set of cross-faults. Here we use sub-pixel correlation of optical satellite imagery to measure the displacement, finite strain and rotation of the near-field coseismic deformation to understand the kinematics of strain release along the surface ruptures. We find the average off-fault deformation along the mainshock rupture is 34% and is significantly higher along the foreshock rupture (56%) suggesting it is a less structurally developed fault system. Measurements of the 2D dilatational strain along the mainshock rupture show a dependency of the width of inelastic strain with the degree of fault extension and contraction, indicating wider fault zones under extension than under shear. Measurements of the vorticity along the main, dextral rupture show that conjugate sinistral faults are embedded within zones of large clockwise rotations caused by the transition of strain beyond the tips of dextral faults leading to bookshelf kinematics. These rotations and bookshelf slip can explain why faults of different shear senses do not intersect one another and the occurrence of pervasive and mechanically unfavorable cross-faulting in this region. Understanding the causes for the variation of fault-zone widths along surface ruptures has importance for reducing the epistemic uncertainty of probabilistic models of distributed rupture that will in turn provide more precise estimates of the hazard distributed rupture poses to nearby infrastructure.

Propagation of large earthquakes as self-healing pulses or mild cracks.

Abstract: Observations suggest that mature faults host large earthquakes at much lower levels of stress than their expected static strength1–11. Potential explanations are that the faults are quasi-statically strong but experience considerable weakening during earthquakes, or that the faults are persistently weak, for example, because of fluid overpressure. Here we use numerical modelling to examine these competing theories for simulated earthquake ruptures that satisfy the well known observations of 1–10 megapascal stress drops and limited heat production. In that regime, quasi-statically strong but dynamically weak faults mainly host relatively sharp, self-healing pulse-like ruptures, with only a small portion of the fault slipping at a given time, whereas persistently weak faults host milder ruptures with more spread-out slip, which are called crack-like ruptures. We find that the sharper self-healing pulses, which exhibit larger dynamic stress changes compared to their static stress changes, result in much larger radiated energy than that inferred teleseismically for megathrust events12. By contrast, milder crack-like ruptures on persistently weak faults, which produce comparable static and dynamic stress changes, are consistent with the seismological observations. The larger radiated energy of self-healing pulses is similar to the limited regional inferences available for crustal strike-slip faults. Our findings suggest that either large earthquakes rarely propagate as self-healing pulses, with potential differences between tectonic settings, or their radiated energy is substantially underestimated, raising questions about earthquake physics and the expected shaking from large earthquakes. Numerical simulations indicate that seismological observations of large megathrust earthquakes are better matched by crack-like ruptures on persistently weak faults than by self-healing pulse-like ruptures on stronger faults.

Feature-Level Attention-Guided Multitask CNN for Fault Diagnosis and Working Conditions Identification of Rolling Bearing.

Abstract: Accurate and real-time fault diagnosis (FD) and working conditions identification (WCI) are the key to ensuring the safe operation of mechanical systems. We observe that there is a close correlation between the fault condition and the working condition in the vibration signal. Most of the intelligent FD methods only learn some features from the vibration signals and then use them to identify fault categories. They ignore the impact of working conditions on the bearing system, and such a single-task learning method cannot learn the complementary information contained in multiple related tasks. Therefore, this article is devoted to mining richer and complementary globally shared features from vibration signals to complete the FD and WCI of rolling bearings at the same time. To this end, we propose a novel multitask attention convolutional neural network (MTA-CNN) that can automatically give feature-level attention to specific tasks. The MTA-CNN consists of a global feature shared network (GFS-network) for learning globally shared features and K task-specific networks with feature-level attention module (FLA-module). This architecture allows the FLA-module to automatically learn the features of specific tasks from globally shared features, thereby sharing information among different tasks. We evaluated our method on the wheelset bearing data set and motor bearing data set. The results show that our method has a better performance than the state-of-the-art deep learning methods and strongly prove that our multitask learning mechanism can improve the results of each task.

Evolution of a salt-rich transtensional rifted margin, eastern North Pyrenees, France

Abstract: In this field study we reinterpret the narrow eastern North Pyrenean Zone, France, as an inverted salt-rich transtensional rift system based on identification of halokinetic depositional sequences across rift platform to distal rift margin domains with a cumulative throw of >2.8 km on steep Cretaceous faults. The rift platform records extension on detached rotational faults above Triassic evaporites from Jurassic to Aptian with uplift and erosion during the Albian. Transtensional Aptian–Albian minibasins align along the salt-rich rift margin fault zone. In the Aptian–Albian main rift large en echelon synclinal minibasins developed between salt walls, although Jurassic diapiric evolution is likely. Upper Cretaceous units locally record continuing diapirism. The Boucheville and Bas Agly depocentres, altered by synrift HT metamorphism, form the distal rift domain terminating south against the North Pyrenean Fault. The narrowness of the Pyrenean rift, shape of minibasins, en echelon oblique synclinal depocentres and folds coupled with a discontinuous distribution and intensity of HT metamorphism support a transtensional regime along the Iberia–Europe plate margin during late Early and early Late Cretaceous. In this model, the distal European margin comprises deep faults limiting laterally discontinuous crustal domains and ‘hot’ pull-apart basins with mantle rocks directly beneath sedimentary cover. Supplementary material: A table summarizing the stratigraphy of the NE Pyrenees and an interpreted Google Earth view of the Quillan syncline and minibasin are available at https://doi.org/10.6084/m9.figshare.c.5100036

Fault pattern and seismotectonic style of the Campania – Lucania 1980 earthquake (Mw 6.9, Southern Italy): new multidisciplinary constraints

Abstract: New fault trace mapping and structural survey of the active faults outcropping within the epicentral area of the Campania-Lucania 1980 normal fault earthquake (Mw 6.9) are integrated with a revision of pre-existing earthquake data and with an updated interpretation of the CROP 04 near-vertical seismic profile to reconstruct the surface and depth geometry, the kinematics and stress tensor of the seismogenic fault pattern. Three main fault alignments, organized in high-angle en-echelon segments of several kilometers in length, are identified and characterized. The inner and intermediate ones, i.e. Inner Irpinia (InIF) and Irpinia Faults (IF), dip eastward; the outer Antithetic Fault (AFA) dips westward. Both the InIF and the IF strike NW-SE along the northern and central segments and rotate to W-E along the southern segments for at least 16 km. We provide evidence of surface coseismic faulting (up to 1m) not recognised before along the E-W segments and document coseismic ruptures with maximum vertical displacement up to ~1m where already surveyed from other investigators 40 years ago. Fault/slip data from surface data and a new compilation of focal mechanisms (1980-2018) were used for strain and stress analyses to show a coherent NNE-directed least principal stress over time and at different crustal depths, with a crustal-scale deviation from the classic SW-NE tensional direction across the Apennines of Italy. The continuation at depth of the outcropping faults is analysed along the trace of the CROP-04 profile and with available hypocentral distributions. Integrating all information, a 3D seismotectonic model, extrapolated to the base of the seismogenic layer, is built. It outlines a graben-like structure with a southern E-W bend developed at depth shallower than 10-12 km, at the hanging wall of an extensional NE- to E-dipping extensional basal detachment. In our interpretation, such a configuration implies a control in the stress transfer during the 1980 earthquake ruptures and provides a new interpretation of the second sub-event, occurred at 20 seconds. Our reconstruction suggests that the latter ruptured a hanging-wall NNE-dipping splay of the E-W striking main fault segment and possibly also an antithetic SSW-dipping splay, in two in sequence episodes.