Uncertainty-Aware, Structure-Preserving Machine Learning Approach for Domain Shift Detection From Nonlinear Dynamic Responses of Structural Systems

In this paper, a structural health monitoring (SHM) system based on multi-scale cross-sample entropy (MSCE) is proposed for detecting damage locations in multi-bay three-dimensional structures. The location of damage is evaluated for each bay through MSCE analysis by examining the degree of dissimilarity between the response signals of vertically-adjacent floors. Subsequently, the results are quantified using the damage index (DI). The performance of the proposed SHM system was determined in this study by performing a finite element analysis of a multi-bay seven-story structure. The derived results revealed that the SHM system successfully detected the damaged floors and their respective directions for several cases. The proposed system provides a preliminary assessment of which bay has been more severely affected. Thus, the effectiveness and high potential of the SHM system for locating damage in large and complex structures rapidly and at low cost are demonstrated.

The Structural Health Monitoring (SHM) system is a method for evaluating and monitoring the integrity of the structure. It has been widely used in various engineering sectors, such as in aerospace, civil and energy sectors due to its ability to react to structural changes on an online basis or in real-time monitoring. The SHM system was able to evolve and work in a variety of structures or components due to the rapid development of sensor technology. It is believed that SHM will become an important tool in various industries for structural monitoring in future years. The paper presents a summary of the latest SHM technology in civil and aviation technology applications as well as the challenges in cost analysis and certification issues for the implementation of SHM.

The continuous Non-Destructive Evaluation of assets for long-term assurance of performance has led to several developments over the deployment of a Real-Time Structural Health Monitoring (SHM) system. Considering the challenges involved under the implementation of an SHM system for the applications working under harsh environmental conditions with limited access to power sources this work is aimed to contribute towards overcoming those challenges by using the noise from the structure’s machinery or any ambient source as an alternative energy source and employing Fiber Optics based sensing, for its applicability under harsh environments. The required SHM system is realized with the cross-correlation of a fully diffused noise field, sensed using the Fiber Bragg Grating (FBG) sensors at two random locations. With no control on the input received as noise, to this end, a method is developed based on a Deep Learning framework, which is aimed towards a Universal Deployment of the passive SHM system. The methodology is designed to perform the health monitoring of the system, independent of the input perturbations. The validation performed on simulation data has demonstrated the feasibility of the developed technique towards the required kind of passive SHM system.

Composite aircraft structures commonly use bonded joints to adhere the wing and fuselage skins to the wing spars, stringers, ribs, and frames. Unfortunately the bond out-ofplane normal (peel) strength is many orders of magnitude less than the fiber strength of the adjacent laminates. Moreover, the adhesive failure is brittle in nature so that it is difficult to monitor an impending overload or fatigue failure. In order to avoid failures due to interlaminar damage without using high margins of safety, a Structural Health Monitoring (SHM) system can be used. While extensive literature on SHM systems for fiber-driven damage exists, there is no mention of an SHM system that specifically addresses these types of matrix-driven damage. The proposed SHM system spans the acquisition, location, reconstruction, and damage type decision of Acoustic Emission (AE) damage signals. An algorithm is presented to calculate the location of AE damage signals in a laminated composite structure using an over-determined approach to minimize location errors. Also, a novel procedure is presented to simultaneously reconstruct the loading source time series for a recorded damage signal and identify the source type. Parametric studies are presented for both algorithms to gauge their effectiveness under varying levels of noise, number of sensors, material properties and different types of uncertainty. Two experimental studies were undertaken to exercise the SHM algorithm presented in this study. The first experiment models a predominantly single type quasi-static damage case, and the second experiment studies a mixed damage impact case.

Primary and secondary aircraft structures are commonly made from laminated composite materials. Unfortunately, the interlaminar tensile and shear strengths are many orders of magnitude less than the in-plane strengths because they are driven by the strength of the matrix as opposed to that of the fibers. Moreover, the matrix strength is brittle in nature so that it is difficult to monitor an impending overload or fatigue failure. In order to avoid failures due to interlaminar damage without using high margins of safety, a Structural Health Monitoring (SHM) system can be used. While extensive literature on SHM systems for fiber-driven damage exists, there is no mention of an SHM system that specifically addresses these types of matrix-driven damage. The proposed SHM system spans the acquisition, location, reconstruction, and damage type decision of Acoustic Emission (AE) damage signals. An algorithm is presented to calculate the location of AE damage signals in a laminated composite structure using an over-determined approach to minimize location errors. Also, a novel procedure is presented to simultaneously reconstruct the loading source time series for a recorded damage signal and identify the source type. An example analytical study is presented to calculate the effects of uncertainty on the location and damage type predictions. Two experimental studies were undertaken to exercise the SHM algorithm presented in this study. The first experiment studies a mixed damage blunt impact case on a flat composite plate for different impact energies. The second experiment studies blunt impact of a reinforced composite plate that models a composite wing skin with bonded stringers and ribs.

Structure damage modelling for the simulation of elastic wave propagation is essential for simulation-based testing of structural health monitoring (SHM) systems. Reflection, transmission and other complex phenomena related to elastic wave interaction with structure defects need to be reproduced by the simulation method to replace physical testing by numerical one and aid the design process of SHM system. Authors review commonly used methods for the simulation of elastic wave propagation pointing out their advantages and disadvantages. Subsequently methods for crack and delamination modelling are presented. Proposed models are used for numerical testing of designed SHM system. Simulation results are compared with experimental ones, showing very good agreement. Finally, structure damage models are used for testing delamination localization algorithm.

Structural health monitoring (SHM) is the process of implementing a damage identification strategy which can be utilized in several applications including aerospace, civil and mechanical engineering infrastructure. Damage is defined as changes to the material and/or geometric properties of these systems. These changes adversely affect the current or future performance of the system. In order to identify damage in a suitable and meaningful manner, the damaged state is compared with other usually undamaged states. This study focuses on a structural health monitoring (SHM) system based on detecting shifts in natural frequencies of the structure. This structural health monitoring system incorporates a low power wireless transmitter that sends a warning signal when damage is detected in a structure. The damage detection technique is implemented on composite structures which are widely used in many applications including aeronautical and aerospace. An automated damage detection system capable of providing information of damage locations based on the finite element analysis and able to compare damage events to other historical data is also proposed in this paper and initially implemented using a microcontroller chip. Moreover, a control methodology using piezoelectric fiber composites, such as active fiber composites (AFCs) and microfiber composites (MFCs), is included as part of the system for vibration suppression purposes. The advantages of using piezoelectric fiber composite actuators are their high performance, flexibility, and durability when compared with the traditional piezoceramic (PZT) actuators. The proposed system may be implemented in many structural components such as aircraft frames and bridges. This SHM technology may help replace the current time-based maintenance scheme with a condition-based one. The condition-based maintenance scheme relies on the ability to monitor the condition of the system and supply information of damage detection to allow a corrective action to be taken.

Objectives: The main objective of this paper to review the various currently available Structural Health Monitoring (SHM) systems for the Polymer Composites particularly in aviation industries. Methods: This paper provides a complete overview about the various types of damage detection technique currently used for the polymer composite structure and also highlights the concept behind the each technique. A brief overview of the research work on experimental and theoretical studies on the various SHM systems is considered and several research papers on these topics are cited. This paper reviews various recent advancement made in the field of damage monitoring techniques, particularly the Piezoresistivity technique with the advent of nanomaterials. Outcomes: This paper serves as an effective source of literature for those interested in pursuing research in the Structural Health Monitoring (SHM) System for the polymer Composites. This paper also provides an effective source of information about the possible research gap in the field the SHM.

Recent structural failures such as the I-35W Mississippi River Bridge in Minnesota have underscored the urgent need for improved methods and procedures for evaluating our aging transportation infrastructure. This research seeks to develop a basis for a Structural Health Monitoring (SHM) system to provide quantitative information related to the structural integrity of metallic structures to make appropriate management decisions and ensuring public safety. This research employs advanced structural analysis and nondestructive testing (NDT) methods for an accurate fatigue analysis. Metal railroad bridges in New Mexico will be the focus since many of these structures are over 100 years old and classified as fracture-critical. The term fracture-critical indicates that failure of a single component may result in complete collapse of the structure such as the one experienced by the I-35W Bridge. Failure may originate from sources such as loss of section due to corrosion or cracking caused by fatigue loading. Because standard inspection practice is primarily visual, these types of defects can go undetected due to oversight, lack of access to critical areas, or, in riveted members, hidden defects that are beneath fasteners or connection angles. Another issue is that it is difficult to determine the fatigue damage that a structure has experienced and the rate at which damage is accumulating due to uncertain history and load distribution in supporting members. A SHM system has several advantages that can overcome these limitations. SHM allows critical areas of the structure to be monitored more quantitatively under actual loading. The research needed to apply SHM to metallic structures was performed and a case study was carried out to show the potential of SHM-driven fatigue evaluation to assess the condition of critical transportation infrastructure and to guide inspectors to potential problem areas. This project combines the expertise in transportation infrastructure at New Mexico State University with the expertise at Sandia National Laboratories in the emerging field of SHM.

The strategic and monetary value of the civil infrastructure worldwide necessitates the development of structural health monitoring (SHM) systems that can accurately monitor structural response due to real-time loading conditions, detect damage in the structure, and report the location and nature of this damage. In the last decade, extensive research has been carried out for developing vibration-based damage detection algorithms that can relate structural dynamics changes to damage occurrence in a structure. In the mean time, the wavelet transform (WT), a signal processing technique based on a windowing approach of dilated ‘scaled’ and shifted wavelets, is being applied to a broad range of engineering applications. Wavelet transform has proven its ability to overcome many of the limitations of the widely used Fourier transform (FT); hence, it has gained popularity as an efficient means of signal processing in SHM systems. This increasing interest in WT for SHM in diverse applications motivates the authors to write an exposition on the current WT technologies. This article presents a utilitarian view of WT and its technologies. By reviewing the state-of-the-art in WT for SHM, the article discusses specific needs of SHM addressed by WT, classifies WT for damage detection into various fields, and describes features unique to WT that lends itself to SHM. The ultimate intent of this article is to provide the readers with a background on the various aspects of WT that might appeal to their need and sector of interest in SHM. Additionally, the comprehensive literature review that comprises this study will provide the interested reader a focused search to investigate using wavelets in SHM.

Near real-time tracking of dynamic properties for standalone structural health monitoring systems

Structural health monitoring (SHM) systems collect and analyze data recorded from civil infrastructure to detect changes in the structural condition and to predict potential damage. However, sensor faults in SHM systems may lead to erroneous structural assessments, thus to inefficient maintenance operations, which could jeopardize the integrity of the structures. Artificial intelligence (AI) algorithms based on analytical redundancy have proven being effective foundations to implement sensor fault diagnosis but are computationally expensive and require large amounts of raw sensor data to be transmitted to centralized servers. Moreover, operating with centralized servers increases the likelihood of single points of failure, which compromises the robustness of SHM systems. To overcome the limitations and to enhance the robustness of SHM systems, this paper proposes a decentralized approach towards fault-tolerant SHM systems based on the concept of Artificial Intelligence of Things (AIoT). Fault diagnosis algorithms based on AI are embedded into sensor nodes deployed for SHM, using Internet of Things technologies to be computed at the edge of civil infrastructure, avoiding large amounts of data to be sent to central servers. In addition, the sensor nodes embed algorithms that collect, process, and analyze the sensor data to assess the condition of the structures being monitored. The result is a smart (i.e., decentralized, fault-tolerant, and AIoT-based) SHM system that is calibrated in a laboratory experiment and then deployed on a bridge for field validation. It is expected that the approach presented in this paper will improve the robustness of SHM systems in assessing the structural condition of civil infrastructure.

Structural Health Monitoring (SHM) systems are widely regarded as capable of significantly reducing inspection costs of safety-critical structures in industries such as aerospace, nuclear, and oil and gas, among others. Successful SHM systems can be considered those which combine good sensitivity to defects, preferably with the capability of localization and identification, with a low sensor density. Techniques based on sparse arrays of sensors which generate and receive guided waves are among the most promising candidates. Guided waves propagate over large distances and certain modes have the ability to transmit through a variety of structural features leading to a relatively small number of distributed sensors being able to cover the structure. In complex structures, which contain high densities of structural elements, the timetraces obtained are often too complex to be directly interpreted due to the large number of overlapping reflections. In this case, the Baseline Subtraction technique becomes attractive. In this method a current signal from the structure is subtracted from a signal which has been acquired during the initial stages of operation of the structure. This eliminates the need for interpretation of the complex raw time signal and any defects will be clearly seen provided the amplitude of the residual signal obtained after subtraction of the baseline signal is sufficiently low when the structure is undamaged. However, it is well known that environmental effects such as stress, ambient temperature variations and liquid loading affect the velocity of guided waves; this modifies the time-traces and leads to high levels of residual signal if a single baseline, taken under different conditions, is used. Of these effects, temperature variations are the most commonly encountered and are critical since they affect not only the wave propagation but also the response of transducers. The present work aims to demonstrate the potential of guided wave health monitoring of large area complex structures. It starts with a general literature review on inspection and monitoring of large area structures, in which the advantages and disadvantages of this technique compared to other well-established SHM techniques 2 are presented. The design and behaviour of two different temperature-stable transducers generating high A0 or S0 mode purity in the sub-200kHz frequency region are described. The efficiency of different signal processing techniques aimed at reducing or eliminating the influence of temperature on wave propagation is evaluated and a temperature compensation signal processing strategy is proposed. Finally, a large metallic structure is used to demonstrate a sparse-array SHM system based on this signal processing strategy, and imaging algorithms are used to combine the information from a large number of sensor combinations, ultimately leading to the localization of defects artificially introduced in the structure.

Fault diagnosis (FD), comprising fault detection, isolation, identification and accommodation, enables structural health monitoring (SHM) systems to operate reliably by allowing timely rectification of sensor faults that may cause data corruption or loss. Although sensor fault identification is scarce in FD of SHM systems, recent FD methods have included fault identification assuming one sensor fault at a time. However, real-world SHM systems may include combined faults that simultaneously affect individual sensors. This paper presents a methodology for identifying combined sensor faults occurring simultaneously in individual sensors. To improve the quality of FD and comprehend the causes leading to sensor faults, the identification of combined sensor faults (ICSF) methodology is based on a formal classification of the types of combined sensor faults. Specifically, the ICSF methodology builds upon long short-term memory (LSTM) networks, i.e. a type of recurrent neural networks, used for classifying ‘sequences’, such as sets of acceleration measurements. The ICSF methodology is validated using real-world acceleration measurements from an SHM system installed on a bridge, demonstrating the capability of the LSTM networks in identifying combined sensor faults, thus improving the quality of FD in SHM systems. Future research aims to decentralize the ICSF methodology and to reformulate the classification models in a mathematical form with an explanation interface, using explainable artificial intelligence, for increased transparency.

In recent years, the development of quick and streamlined methods for the detection and localization of structural damage has been achieved by analysing key dynamic parameters before and after significant events or as a result of aging. Many Structural Health Monitoring (SHM) systems rely on the relationship between occurred damage and variations in eigenfrequencies. While it is acknowledged that damage can affect eigenfrequencies, the reverse is not necessarily true, particularly for minor frequency variations. Thus, reducing false positives is essential for the effectiveness of SHM systems. The aim of this paper is to identify scenarios where observed changes in eigenfrequencies are not caused by structural damage, but rather by non-stationary combinations of input and system response (e.g., wind effects, traffic vibrations), or by stochastic variations in mass, damping, and stiffness (e.g., environmental variations). To achieve this, statistical variations of thresholds were established to separate linear non-stationary behaviour from nonlinear structural behaviour. The Duffing oscillator was employed in this study to perform various nonlinear analyses via Monte Carlo simulations.