European Workshop on Structural Health Monitoring

The use of Lamb waves for the determination of component states and damage detection in thin-walled structures is a generally accepted method in today’s Structural Health Monitoring. Although the behavior of Lamb waves has been dealt with in depth, there are still topics where research is needed, for example, the influence of prestress on Lamb wave propagation. In order to gain profound insight into this area, experimental methods and results on wave propagation under the influence of prestress have been presented in previous work by the authors. These results are characterized by the use of a 2d Fourier transform to evaluate the observation of this behavior over large frequency ranges. Based on these results, the work presented here deals with a 2d numerical modeling approach of the wave propagation in a prestressed structure. The focus is on analyzing the suitability of linear and non-linear material models for reproducing the wave velocity changes. For this purpose, the numerically obtained results for different material models are compared with experimental data. The numerical data is generated using a FE model which combines a nonlinear static computation for the prestress and a linear eigenfrequency computation for the dispersion behaviour of Lamb waves. Finally, the comparison of simulated and experimental data will provide a well-founded basis for further numerical analysis on this topic.

In this paper results of damage assessment of composite panel using guided wave propagation phenomenon are presented. Elastic waves excitation is based on piezoelectric transducer (PZT) and air-coupled transducer (ACT) while the waves sensing is based on scanning laser Doppler vibrometry. It thus forms the full non-contact diagnostic approach. Thin panel made of fibre-reinforced polymer is investigated. The problem of optimal slope angle of ACT and the possibility of symmetric and antisymmetric elastic wave modes excitation is investigated. Contact (PZT) and non-contact (ACT) elastic waves generation methods and their influence on the artificial damage (Teflon inserts) localization results are compared. Moreover, the influence of single and multiple acoustic wave sources on artificial damage localization results and the problem of panel coverage area by elastic waves with large amplitudes to improve damage sensitivity are investigated. Different locations of ACTs and their influence on damage detection results are investigated. Two damage imaging algorithms based on elastic waves root mean square (RMS) energy maps and wavefield irregularity mapping (WIM) have been proposed.

In structural health monitoring (SHM) a considerable amount of damage detection algorithms based on guided waves (GW) have been developed. Most of them rely on extensive transducer networks, besides preliminary reference measurements of the structures. This originated a growing demand for hardware simplification and cost reduction of the wave-based SHM methods, driving the conception of new solutions enabling both: the reduction in the amount of sensors required for doing measurements, as well as a diminution of quantity of signals needed for the algorithms to work. The simplification in damage detection procedures can be achieved by using a novel type of special shaped frequency steerable acoustic transducers (FSATs). The spiral shape of these FSATs allows focusing wave energy in a certain direction, which is associated with their excitation frequency. Thanks to this property, presence of damage can be established by identifying signal reflections, while its localization can be determined based on time of flight and the relationship between direction of propagation and its spectral content. This article presents the concept of baseline-free damage detection using FSATs over an aluminium plate with point damage through Finite Element (FE) analysis. Numerical simulations were performed for several cases, varying excitation frequency and damage position.

The paper is focused on the investigation of Lamb wave signals measured in different operational and environmental conditions of an ultralight aircraft. Transducers were mounted on the right side of the fuselage skin behind the pilot seat and the circular cover located on the bottom of the fuselage. The fatigue crack of about 1 mm in length was cycled into the drilled hole in the center of the cover part in the laboratory before the test. The side fuselage skin remained intact. Five test flights were conducted within four months. The temperature effect on Lamb wave signals was realized through measurements in different flight altitudes ranging between 2,000 and 7,000 ft. Both A0 and S0 mode were analyzed within the frequency range of 50 to 500 kHz. Based on the collected data, it can be concluded that signals influenced by environmental and operational conditions and signals affected by the damage can be easily distinguished.

Damage localization is one of the most challenging topics within Structural Health Monitoring (SHM) in aeronautics, especially when the structure is manufactured out of carbon fiber-reinforced composite materials. Using ultrasonic guided waves (particularly Lamb waves), generated and recorded with piezoelectric transducers, is also challenging in this type of material. Otherwise, traditional methods used for this task are subjected to physics-based knowledge of the problem, such as damage imaging algorithms like delay-and-sum and RAPID. This paper presents an entirely data-driven approach, based on the ability of Deep Learning (DL) techniques (particularly those based on Convolutional Neural Networks – CNNs –) to extract features of interest for damage imaging from a pre-dataset. In this work, the selected feature to be estimated is the normal distance from the propagation path of the guided wave to a simulated damage, which allows, in combination with an especially designed positioning algorithm, to locate with high accuracy defects, even in different positions than the used for the training of the network (a fixed grid of points over the analysis zone). This paper presents the application of the method to a real composite material specimen, as well as the recorded results obtained from additional datasets recorded with the simulated damage (a piece of blu-tack) attached to different random positions other than those of the training grid.

The difference between damage-affected signals and temperature-affected signals is investigated. Temperature effects were examined by measurements in a climatic chamber. Sensors were mounted on the skin of a CFRP sandwich structure which was exposed to temperatures from −50 to +50 °C with the step of 10 ℃. Both A0 and S0 modes were analyzed in the frequency range of 50–300 kHz. Time of flight delays indicating a velocity decrease with increasing temperature were confirmed for the corresponding modes. However, with decreasing temperature, frequency shifts towards higher frequencies were observed, which caused difficulties for dispersion curve determination. Damage–type signals were demonstrated by impact-induced damage. Time of flight change was examined in relation to the temperature and damage. Temperature affected time delay/time advance increased with the signal length while damage influenced only some part of the signal. These findings enabled to distinguish the cause of the time of arrival change. Therefore, false alarms regarding damage detection are likely to be reduced.

Monitoring and inspection of pipes are paramount in several industry sectors. A convenient approach employs shear horizontal ultrasonic guided waves that propagate circumferentially around the pipe, known as CSH waves. Typically, a pair of transducers, a transmitter and a receiver, can be used to interrogate the whole circumference. Conventional shear horizontal transducers generate waves that propagate in both directions, namely, clockwise and counterclockwise. Bidirectional generation complicates the interpretation of the received signal since the scattered waves from a potential defect can mix themselves or with one of the two direct waves generated by the transmitter. In this paper, we address how unidirectional generation can ease the signal interpretation task for circumferential guided wave testing utilizing finite element simulations. Unidirectional generation is based on dual periodic elements transducer driven by phased pulses, following a design previously presented. Generation and reception were performed with separate transducers on the outer surface of a 324 mm outer diameter, 6.25 mm wall thickness steel pipe. Corrosion-like defects were modelled on the inner surface. We show that, when the angular position of the defect approaches 180° from the transmitter, the defect echo can be almost completely masked by the wave generated in the opposite direction. Unidirectional generation of CSH waves proves itself to be an important feature in pipeline integrity monitoring, providing more reliable signal interpretation.

The guided-wave-based damage detection methodology is a well-established branch of Structural Health Monitoring. There are, however, many different approaches to processing and interpretation of signals acquired using guided waves. This paper presents the results of the SYPIN project in which a large database acquired in a diverse set of experiments was assembled. The experiments included static and variable environmental conditions, composite and metallic specimens of simple and complex geometries. The set includes, in particular, results of experiments on real-life aircraft specimens. This dataset was then used to benchmark several damage detection and quantification methodologies, including various classifiers and feature extraction methods. The performance of the methods is evaluated in a series of experiments, including knowledge transfer between different specimens. The paper presents the results of this research and provides a brief description of the dataset, which is made available for the scientific community’s use.

Asset inspection of large structures such as storage tanks in the oil, gas and petrochemical industry is a challenging and time-consuming process. Compared to the state-of-the-art inspection methods of these structures using ultrasonic bulk wave phased array techniques, guided waves provide a mechanism for monitoring inaccessible areas and increasing the testing range to speed up the inspection process. Combining these two approaches, an efficient inspection scheme can be achieved. For this purpose, in this work, the possibility of exciting a single higher order guided wave mode at a low dispersion region is examined, using a conventional linear phased array on a 10 mm thick sample. An analytical model based on modal analysis is derived. Then, the time delays across each element of the array are optimally selected and a necessary condition on the pitch is provided, to enhance the purity of the desired mode. The results are promising, illustrating that single mode excitation is possible at high frequency-thickness products, greater or equal than 15 MHz $$\cdot $$ · mm. Furthermore, the phased array transducer allows for dynamic mode selection capabilities; that is, feeding the appropriate excitation signals on each channel, different modes can be excited, without having to physically alter the transducer configuration.

A multiresolution finite wavelet domain method, that utilizes Daubechies wavelet and scaling functions for the hierarchical approximation of state variables, is presented. The multiresolution approximation yields a hierarchical set of equations of motion involving the coarse component of generalized displacements, while additional equations of finer components are subsequently added. A coarse solution is first calculated, and finer solutions can be sequentially superimposed on the coarse solution until convergence to the final solution is achieved. Moreover, it is shown that each resolution can model specific bandwidths of wavenumbers, thus providing a unique capability to separate coexisting wave modes and detect converted and reflected waves in the presence of damage. Two wavelet-based beam elements are explored, the first encompasses the Timoshenko shear beam theory and the second a high-order layerwise laminate theory for the accurate prediction of both symmetric and antisymmetric guided waves. Numerical results illustrate the inherent property of the method to a priori localize and isolate coexisting guided wave modes and their conversions, induced by different material regions and weak or debonded layer interfaces, thus demonstrating the method’s intrinsic capabilities towards the design of wave-based SHM systems.

Acoustic-ultrasound (AU) sensing is an effective, and powerful tool for the non-destructive testing and evaluation of composite and metallic material structures. The AU technology consists of sending and receiving low frequency acoustic pulses at a predetermined angle of incidence into a material under inspection. Analysis of the detected acoustic-ultrasound waveform characteristics provides a clear representation of the mechanical state of the structure.Our research group at Redondo Optics is currently in the process of testing and demonstrating the performance demonstration of a unique fully integrated, miniature, lightweight, and power efficient, hybrid “piezoelectric/fiber optic” acousto-ultrasound sensor network (FAULSense™) SHM system suitable for the real-time, in-situ, and un-attended detection, identification, localization, classification, of static and dynamic load environments representative of the structural state of large rotorcraft structures. The hybrid PZT/FBG FAULSense™ SHM system is a highly integrated, ultra-sensitive, high frequency response (DC to 5-MHz), acousto-ultrasound structural health monitoring (SHM) system that uses a wide area coverage flexible and conformal strip-patch sensor network integrating a stack of thin-film piezoelectric actuators used to input control structural excitation of acoustic signals over the entire, or discrete locations, of an aircraft structure (fuselage, wings, tails, rotor blades, etc.), and a distributed array of fiber Bragg gratings (FBG) acousto-ultrasound sensors used to detect and measure the corresponding waveform structural response to the PZT excitation signals, that contains structural information representative of potential damage associated with excessive loads, fatigue, impacts, fractures, delamination, corrosion, radiation, and structural damage of large composite and metallic structures commonly found in military rotorcraft.

The guided wave (GW) based structural health monitoring method has been a promising method because this method can cover a wide monitoring range and it is sensitive to small damage. However, online damage quantification is difficult as damage initiation and growth are affected by various uncertainties. In addition, various time-varying conditions introduce uncertainty effects on GW features, resulting in a decrease in the accuracy of damage quantification. Therefore, a multi-source feature fusion Gaussian mixture model (MSFF-GMM) is proposed to improve the accuracy of damage quantification. Firstly, MSFF-GMM is constructed by fusing the feature information from multiple channels and multiple specimens under time-varying conditions. Then, the migration index calculated by Kullback-Leibler divergence is used to measure the difference between the GMMs constructed at different damage degrees. Finally, the proposed method is validated for online damage quantification on the typical aircraft lug structure under time-varying dynamic load conditions.

Composite materials are subject to internal damages that can threaten structural integrity while being invisible to the naked eye. StructuralHealth Monitoring (SHM) allows to ensure in real-time that aircraft substructures can still perform their function. Among all the technologies used in SHM, the emission/reception of Lamb waves are very popular. Algorithms using Lamb waves for damage detection and localization already exist in the literature but damage size estimation is still an open problem. In this paper, we propose an approach to quantify delamination damage size by relying on an analytical scattering model. The structure considered is a plate equipped with piezoelectric transducers (PZT) acting both as actuators and sensors. We use the framework of the Mindlin-Kane plate theory to describe S0-mode Lamb wave propagation. The damage considered is a cylindrical inhomogeneity where the mechanical properties are different from the rest of the plate. The analytical model takes into account the signal emission by a PZT, the scattering by the damage, and the reception by a sensor PZT. This model is then used in an identification process to estimate the size of the damage by minimizing a dedicated cost function. The proposed approach is applied on aluminum plate simulation data.

Guided wave-based structural health monitoring techniques require accurate and fast simulation tools for high-frequency wave propagation in the laminate structures. This article develops an accurate and computationally efficient time-domain spectral finite element (SFE) for wave propagation analysis of laminated composite and sandwich beam and panel-type structures based on the efficient layerwise zigzag theory. It considers the axial displacement to follow a global third-order variation with a layerwise linear variation across the thickness. The independent variables are reduced to only three by imposing the interfacial continuity of transverse shear stress and shear traction-free conditions at the top and bottom surfaces. Regardless of the number of layers in the laminate, the element has only four degrees of freedom (DOFs) per node $$u_0$$ u 0 , $$w_0$$ w 0 , $$\frac{dw_0}{dx}$$ d w 0 dx , and $$\psi _0$$ ψ 0 . The deflection $$w_0$$ w 0 is interpolated using the C $$^1$$ 1 -continuous Lobatto basis function, whereas $$u_0$$ u 0 and $$\psi _0$$ ψ 0 employ the C $$^0$$ 0 -continuous Lobatto basis shape functions. A thorough numerical study is accomplished to validate and evaluate the proposed element’s accuracy and efficiency for free vibration and Lamb wave propagation analysis of laminated composite and sandwich panels. The study reveals that the developed element is superior to its conventional counterpart and other existing 1D elements with a similar number of DOFs.

Guided waves are gaining increased interest in SHM, thanks to some distinct advantages. For guided-wave-based localisation strategies, information on the group velocity is required; therefore, determination of accurate dispersion curves is invaluable. However, for complex materials, the wave speed is dependent on the propagation angle. From experimental observations of dispersion curves, measured using a two-dimensional Fourier transform, a system identification procedure can be used to determine the estimated value and distribution for the governing material properties. Markov-chain Monte Carlo (MCMC) sampling can provide a way of simulating samples from these distributions, which would require solving dispersion curves many times. By using a novel Legendre polynomial expansion approach, the computational cost of dispersion curve solutions is greatly reduced, making the MCMC procedure a more practical approach In this work, a scanning laser Doppler vibrometer is used to record the propagation of Lamb waves in a carbon-fibre-composite plates, and points on the dispersion curve are extracted. These observations are then fed into the MCMC material identification procedure to provide a Bayesian approach to determining properties governing Lamb wave propagation at various angles in the plate. The distribution of parameters at each angle is discussed, including the inferred confidence in the predicted parameters.

Structural damage can result in observable changes of the signal acquired by network of PZT sensors, due to elastic wave interaction with damage. There are two approaches how to utilize PZT sensors for SHM purposes. One of the approaches follows closely classical ultrasonic testing. In that case, short-pulse excitation of PZT transducers is used, thus guided wave packets can be scattered on different elements of structure, eventually also on the damage itself. The disturbance of the scattered wavefield is the basis of damage detection and evaluation. In the different approach harmonic excitation of PZT is used, thus steady elastic waves are excited in the structure. The signal can be gathered in the pulse-echo scheme, i.e., when a single transducer is used both as an actuator and a receiver of waves, as well in the pitch-catch scheme, when a pair of transducers is used. An approach for damage detection with use of a network of PZT sensors excited with harmonic signals in the pitch-catch scheme will be presented and its properties and damage detection capabilities will be discussed. In addition method for data classification based on the Bayesian approach will be demonstrated and compared to other approaches to data classification.

Ultrasonic SHM approaches using bulk or guided waves has been well documented. Here the waves are generate using PZT transduction that requires excitation inputs using either high voltage or high current or both. Several critical applications such as component in explosion susceptible environment, extreme temperature condition, etc. have limitations in deployment of high voltage/current excitation. In this paper, the use of wave-guide-based excitation is explored for plate and pipe like geometries. The waveguide help isolate the excitation part of the ultrasound generation, away from the component that may be in a hostile condition. The ultrasound wave modes excited in the waveguide is coupled to the structure and optimized to generate the desired guided wave modes in the geometry of the structures using Finite Element Wave Propagation models. The reception of these modes are also explored using similar ultrasonic waveguides. The detection of defects, in an SHM mode, using this remote generation approach is also demonstrated on a plate geometry component. This article demonstrates the detection of fundamental symmetric wave mode (S0), shear horizontal waves (SH0) and antisymmetric wave mode (A0) in an steel plate by placing a thin wire-like stainless steel waveguide sensor. In order to uniquely identify the three-fundamental plate-guided modes, we map the experimentally measured group velocities as detected by the waveguide sensor to theoretically obtained group velocity dispersion curves. The analytically obtained FEM results are evaluated experimentally using time-of-flight measurements. In future, this sensor arrangement may be utilized in areas like high-temperature furnaces, engines and aircrafts for detecting cracks.

Fiber optical grating (FBG) sensors offer several advantages such as small size, ability to be embedded, and insensitivity to magnetic and electric fields, and hence are widely used for monitoring in several sectors. FBG sensors are most commonly used for strain-based structural health monitoring (SHM), but recent developments are also focused on using them for guided waves (GW) based SHM. The coupling of the GW to the fiber is a very complex phenomenon and needs to be studied in detail. The numerical studies with optical fibers are computationally expensive due to the small mesh size needed for capturing the wave propagation. Also, the conversion of the strain profile from the numerical model to the spectral response of the FBG sensor is a non-trivial process needing complex matrix manipulations. Thus, in this paper, a tool for the visualization of the FBG spectrum change subjected to GW is developed. A GUI based tool is developed and validated with experiments to allow easy visualization of the FBG response by the engineers. Through the use of the tool, the fiber need not be modeled thus making the numerical analysis less demanding.

The boom is the key load-bearing component of the crane, and its health seriously threatens the service performance of the crane. In order to ensure the safe production of cranes, nondestructive testing (NDT) and structural health monitoring (SHM) of boom become more and more urgent and important. In this paper, the intelligent defect location algorithm based on helical guided waves is applied to the weld defect detection of U-shaped boom. Intelligent defect location algorithm is an imaging algorithm that combines ellipse imaging principle, evolutionary algorithm and K-means clustering algorithm. This paper verifies the effectiveness of this algorithm for weld defect location of U-shaped boom through experimental research. Firstly, the propagation regularity of helical guided waves in the U-shaped boom structure is studied. In addition, the elliptical imaging algorithm is used to image and analyze the weld defects in the U-shaped boom, and the location of the defects is preliminarily determined. The defect location analysis of weld defects is carried out by using the intelligent defect location algorithm. By comparing the imaging results of the two methods, it is found that the intelligent defect location algorithm has higher resolution and can effectively improve the defect location accuracy. This algorithm provides a tool for health monitoring of special-shaped structures based on guided waves.

Due to its many advantages and fast development, additive manufacturing (AM) is evolving from prototyping and simple applications to the manufacturing of complex industrial components. Modern industries including those with critical applications, like aerospace and civil constructions, have recently started adopting AM parts within their structures. This imposes the need to establish reliable and robust techniques to assess and monitor the structural integrity of such components. This study focuses on understanding the propagation of ultrasonic Lamb waves (LWs) in 3D printed poly(lactic acid) (PLA) plates of various infill densities; further, on determining the sensitivity of the propagating wave to buried defects within the plates under study. Intact and damaged PLA plates, printed using fused deposition modeling (FDM) with different infill densities, were used for the investigation. LWs were excited using lead-zirconate-titanate (PZT) wafers, and their propagation was visualized through area-scans using a laser Doppler vibrometer. The fundamental LW modes were characterized and successfully used in the detection and localization of damage regions. The obtained results show a high potential of LW-based techniques for the structural health monitoring of 3D-printed structures.

Metal dendrites can form on the electrode surface when a battery is repeatedly charged and discharged, this often results in reduced electrochemical performance and occasionally even internal short-circuiting. Hence, intensive research to study the underlying growth kinetics of dendrites in batteries has been initiated. However, existing technologies such as scanning electron microscopy and x-ray computed tomography are prohibitively expensive for in-operando monitoring. Recently, ultrasound-based techniques have become a promising alternative method for monitoring the health and status of battery cells. However, most of the reported ultrasonic methods utilise bulk waves which propagate through a laminated structure of cathode, anode, and electrolyte. In this study we use miniature transducers to excite fundamental shear-horizontal mode (SH0*) guided ultrasonic waves that propagate along an individual cathode and anode of an aqueous zinc ion battery. Coupled with optical microscopy, the cyclical variations of wave velocity and attenuation can be associated with the zinc strip-ping/plating phenomena that occur at the electrode surface. This demonstrates the sensitivity of the SH0* mode guided wave to dendrite formation at the electrode-electrolyte interface. More importantly, the findings suggest that measurements of guided ultrasonic waves may be used to quantitatively track dendrite growth in batteries.

We investigate the propagation of nonlinear guided waves in plates in view of their application to the identification of the state of stress. The study is performed modelling via Finite Elements the propagation of Lamb waves in a prestressed plate, whose equations of motion are a second-order approximation accounting for both geometric and material nonlinearities. Different intial prestress conditions are investigated, parallel and orthogonal to the direction of wave propagation. A couple of physical phenomena presenting remarkable sensitivity to initial stress emerged using the resonant couple of S1-S2 modes. First, the increase of the secondary wave amplitude with propagation distance is clearly observed together with the sensitivity of the second-harmonic amplitude to the initial state of prestress. Second, a subharmonic and superharmonics are observed, due to the interaction between primary and secondary waves.

Composite structures have been increasingly adopted for designing high performance airframes. However, the critical fracture mechanics prevents a smooth integration of composites in safety designed aircraft. Some compensation strategies are needed to face the occurrence of barely visible damage, including interply delamination and disbonding of stringers adopted to stiffen the airframe. The latter evidences the necessity of crack stoppers in form of rivets or similar connections between the skin and the stringer although composite allow one-piece manufacturing. This brings to light the necessity of crack identification, localization and size estimation to achieve the stringer monitoring and make use of less connections possible. To this end, a novel and promising approach based on guided wave propagation and scattering in composite media is investigated thoroughly using finite element modeling. Firstly, the potentiality of this approach is highlighted making use of a design formulation applied to a linear sensor network located alongside the stringer. Secondly, the interaction between the wave and the stringer is further investigated to achieve simultaneous inspections by an asymmetric design of the sensor network. Theoretical findings demonstrate the ability of guided waves to achieve a local monitoring of the structural performance. The wave focusing can be achieved by locating the sensors appropriately. Numerical outcomes show how a good level of inspection can be achieved making use of fewer sensors possible.

Advanced carbon fiber composites offer substantial environmental benefits over any other traditional material. They are associated with high mechanical, chemical, and thermal properties in addition to being light in weight. In this work, a numerical model for a carbon/epoxy composite plate is developed to analyze and monitor the curing cycle and degree of cure using Lamb waves. With the help of experimental dynamic mechanical analysis results, the storage moduli for this particular woven CFRP are determined throughout the curing cycle and implemented into COMSOL Multiphysics finite element software to simulate the same mechanical property changes of the composite during the curing cycle, while Lamb waves are excited and sensed via piezoelectric transducers embedded within a thin PTFE film. The same is repeated to test a shortened curing cycle trimmed by one hour compared to the original cycle suggested by the manufacturer. Sensed signals are compared to experimental data for both cycles. Then, cure parameters such as minimum viscosity, gelation, and vitrification are deduced from the analysis of voltage and velocity curves of the first received A0 mode packets. These curing parameters are shown offset by the same time deducted from the cycle, thus proving the effectiveness of this computational model.

We present a numerical and experimental framework for the inspection of double cantilever beam composite coupons that are fabricated by bonding two composite laminates. The method is based on a mode-matching procedure in which the experimental dispersion curves of guided wave modes supported by the coupons are compared with those computed by a numerical model in which the bond strength is progressively updated until a match is achieved. Specifically, guided waves are generated in the coupons in the 0–250 kHz range by direct contact excitation with a wedge beam transducer. Their corresponding velocity field is recorded using infrared laser vibrometry. From the measured velocity arrays, the dispersion curves of several higher order modes are extracted and subsequently used to match a semi-analytical finite element model of the coupons. The updating procedure is based on two steps. In the first step, the dispersion data of a coupon with a known pristine bond are used to determine the orthotropic elastic coefficients of the composite laminae. In the second step, these coefficients are fixed and only the bond equivalent Young’s modulus is updated by using the experimental dispersion curves. Analyses performed on coupons with three different level of seeded contamination indicate that our proposed method can quantify their reduction with good accuracy, which makes the method suitable for the efficient non-destructive inspection of structural components.

Shear horizontal (SH) guided waves are great options for nondestructive testing and structural health monitoring (SHM) applications in plates due to their simple dispersion behavior and long-range inspection capability. Being shear waves, they do not leak into fluid that may be present on the plate surface. Magnetostrictive transducers (MSTs) are relatively simple, cost-effective transducers that can generate and receive SH guided waves. However, modelling of the transducer functionality has been quite limited to date. Therefore, in this study, finite element modeling using the COMSOL Multiphysics software was performed to better understand the effect of transducer design variables and transducer layout on SH guided wave-based SHM. We investigated the effect that the thickness of the adhesive and meander coil liftoff have on MST generation characteristics. Additionally, the benefit of two-sided transducers relative to single-sided transducers for symmetric and antisymmetric wave modes was clarified. The results provide suggestions for MST design optimization and further improvement of its capability for SHM and NDE.

This work aims at presenting the development of an analytical model for Structural Health Monitoring (SHM) to evaluate damages on Single Lap Joints (SLJs). A transient waves technique is proposed for assessing debonding area in 3D aluminium SLJ. This approach is based on the interference of elastic Lamb waves generated by piezoelectric (PZT) sensors attached onto the surface of the thin joint and travelling through the adhesion area. The S0 Lamb mode is used to investigate the integrity of adhesive in bonded zone. Several transient wave packets with different driving frequencies were used to investigate bonded area. The main advantage of the proposed model is the ability to easily evaluate the length and position of a damage. Experimental campaign revealed that for fixed disbond length there is a given driving frequency at which the wave packet is attenuated. The main result is the simplicity of the proposed approach, interesting for future applications (i.e., online SHM on working structures).

This research demonstrates the use of genetic algorithms for damage detection in isotropic rods. The spectral element method and a deep-learning-based surrogate model is utilized for simulating wave propagation in an isotropic cracked rod. The genetic algorithm employs results (“numerical experiment") obtained from the spectral element model and the deep-learning-based surrogate to determine the optimized crack locations and crack depths as output parameters. The objective function used in the genetic algorithm is the mean square error between the response obtained from spectral element model and the deep-learning-based surrogate model.

Ultrasonic methods for damage imaging typically use baseline signals from the undamaged part, which are often affected by real operational conditions and may not always be available. This paper proposes a baseline-free damage imaging algorithm based on a combination of virtual time reversal (VTR) and the Reconstruction Algorithm for Probabilistic Inspection of Damage (RAPID) tomographic technique. VTR is an alternative to traditional time reversal as it reduces the burden of physically back-propagating re-mitted signals by applying signal operations between the transmitted and received waveforms. Spatial VTR-based damage indices were here proposed to enhance defect detection as they do not require the time domain reconstruction of re-emitted signals. Experimental results on damaged aluminium and composite specimens showed that the proposed VTR-based damage indices coupled to the RAPID imaging algorithm were able to localise material flaws with a maximum localisation error of ~6 mm and ~2 mm for aluminium and composite samples, respectively.

In this study, the fundamental torsional wave mode T(0, 1), generated by d35 piezoelectric transducers, is utilized in the detection of localized defects such as pitting corrosion. Numerical models based on the finite element method are built to simulate wave propagation and collect data in the time domain. A curved gradient-index phononic crystal (GRIN-PC) lens is implemented to redirect the plane wave to a single point where a single sensor is sufficient to record the interaction of plane wave with the pipe geometry. Parametric studies are carried out and generalized to determine the trends for different defect size, orientation, and axial positions. Signal difference coefficient (SDC) is selected as a damage index based on the modified dispersion properties and the results of principal components analysis. Since the damage information is focused at a single location, the need for using an array of transducers for sensing is eliminated. With the changes in the dispersion of T(0, 1) caused by the GRIN-PC lens, localized defects can be detected earlier than in a plain pipe, before they reach a significant thickness of the pipe wall.

Electromechanical impedance (EMI) spectroscopy is a widely used methodology in structural health monitoring (SHM) to differentiate pristine from damaged structural conditions. In this work, a novel signal processing approach is presented based on distance-based clustering and semi-supervised tree classifiers. The main result of this contribution is a strongly reduction in the number of test frequencies supporting damage classification.The experimental proof of principle is conducted on an aluminum plate with drilled holes of increasing size. The measurements are recorded with the commercially available handheld device AD5933 that covers a frequency range from 1 to 100 kHz. The evaluation of the conductance data to separate pristine and damaged states is performed using only two test frequencies. Implementing this contribution might result in a lower energy consumption and a longer battery lifetime in many practical applications where EMI spectroscopy measurements are used.

As one of the most relevant components in rotary machinery, ball bearings play an important role in diverse areas. To research bearing health state and remaining useful lifetime, several datasets have been developed. Among these datasets, Case Western Reserve University (CWRU) dataset is the most commonly used for bearing diagnosis. A large variety of approaches are applied on CWRU dataset and generating good even the tendency of perfect results. However, most of these approaches are based on supervised learning approaches and focus on classification of bearing faults. In this contribution, in difference to well-known existing approaches, an unsupervised approach combining autoencoder with k-mean is applied on the CWRU dataset. Firstly, the original data are segmented into proper parts. Segments in time domain are transformed to time-frequency domain by adjusting the window length and window function using Short-Time Fourier Transform (STFT), and an associated spectrogram is generated. Spectrogram features are extracted using autoencoder and clustered using K-mean. Various metrics are used to evaluate the performance of the proposed approach. All metrics values demonstrate that this approach could distinguish CWRU bearing from fault-free state to faulty state. As a new result, the requirement of related training datasets of the other approach is – for fault detection – no longer necessary in the future.

Railway bridges are key assets of a countries’ infrastructure, enabling transport of goods and people through freight and passenger trains. The studied structure is a steel railway bridge subjected to cyclic loading, equipped with 98 Fiber Bragg Gratings. A previous study identified train passages as main drivers of damage, isolated and converted them to fatigue damage. This research aims at predicting this damage through machine learning with available operational data as input (train type, train speed, ...) and adding publicly available data (temperature, humidity, ...). The research uses 4 months’ data of train passages and focuses on passenger trains, as too few freight train passages were recorded. Random Forest regression was selected for its ease of implementation with categorical data and high R-squared score. A model was trained for every sensor point. Additionally, the model classifies sensors based on damage predictability. Finally, the models were used to determine long-term damage caused by different bridge loading scenarios. By fixing a parameter like train type and then randomly sampling from train passages, the remaining train passages until a damage threshold is reached are estimated. By repeating this simulation 1000 times for every scenario, remaining train passages distributions are reached, showing best and worst case estimates.

In the recent years, machine learning algorithms have been widely employed for structural health monitoring applications. As an example, Artificial Neural Networks (ANN) could be useful in giving a precise and complete mapping of damage distribution in a structure, including low-intensity or localized defects, which could be difficult to detect via traditional testing techniques. In this domain, Convolutional Neural Network (CNN) are employed in this work along with one-dimensional refined models based on the Carrera Unified formulation (CUF) for surface strain/displacement based damage detection in composite laminates. A layer-wise kinematic is adopted, while an isotropic damage formulation is implemented. In detail, CUF-based finite element models have been exploited in combination with Monte Carlo simulations for the creation of a dataset of damage scenarios used for the training of the CNN. Therefore, the latter is fed with images of the strain or displacement field in a region of particular interest for each sample, which are subjected to the same boundary conditions. The trained CNN, given the strain/displacement mapping of an unknown structure, is therefore able to detect and classify all the damages within the structure, solving the so-called inverse problem.

The capabilities of most non-destructive testing methods have been combined with some degree of automation in recent years, to enhance data acquisition speed, part coverage and inspection reliability. A plethora of automated or semi-automated inspection systems have been engineered to enable the robotic manipulation of specific types of sensors. Robotic inspection systems are usually operated through off-line programmed tool-paths. This approach works well when an accurate model of the part is available and the robotic inspection takes place in a well-structured environment, where the part position is precisely registered with respect to the robot reference system. However, it makes the inspection setup for each new part very time-consuming and dependent on the skills and experience of the robot programmer. Moreover, the real geometry of a part may significantly deviate from its digital counterpart, resulting in inaccurate tool paths. This work introduces a new approach capable of conferring full autonomy to robotic sensing applications, providing a breakthrough in the state-of-the-art. As a result of this work, fully autonomous single-pass geometric and volumetric inspection of complex parts, using one single robotized sensor, becomes possible. This concept can find wide applicability to the open problems of structural health monitoring of the modern age.

The current approach for detecting anomalies in acceleration signals relies extensively on feature engineering. Indeed, detecting rotor imbalances in wind turbines starts by first isolating and then assessing the energy of the 1P harmonic, leading to a feature that is efficient but not failure mode agnostic. While different engineered features can be used concurrently, some anomalies in the acceleration signal might remain undetected by the algorithm, even though they are visually noticeable to a human in the signal’s spectrogram. Thus, this project aims to build an AI algorithm capable of detecting anomalies in spectrograms, agnostic of their origin, providing an early warning for potential structural issues. The proposed algorithm infers spectrograms of acceleration signals through a deep autoencoder. Anomalies are identified based on a custom reconstruction error. A sensitivity analysis is performed for two types of anomaly, in which waveforms with different energy levels are artificially added to an acceleration signal measured from an offshore wind turbine (OWT). For a 1P harmonic anomaly representing 20% of the total signal energy, the proposed approach yielded an efficiency (AUC) equal to 96% thanks to a novel reconstruction error, which significantly increased the performances.

Structural health and condition assessment have become an important part of infrastructural life, due to continuous deterioration caused by nature or human activities. It ensures public safety and saves the economy if carried out properly. Even though deterioration is the natural process of a structure, the rate depends on the functional utility and its maintenance. So, poor maintenance of commercial structures, public buildings, and historical constructions can lead to major damages to sudden collapse. There are several methodologies to predict the health of the structures like destructive, Non-destructive, and sensor-based evaluations, but there is a huge scarce of experienced engineers in interpreting and evaluating the condition of the structures. In this paper, an attempt is made to help engineers in a robust way with fast-moving models to predict the condition at a global level of the structure using Machine Learning for the data obtained from Non-Destructive Tests (NDT) which are in general carried on local level structural elements. In conclusion, a model is proposed which connects between local elemental properties to global behaviour.

Recent advancements in sensor technology, as well as fast progress in internet-based cloud computation; data-driven approaches in structural health monitoring (SHM) are gaining prominence. The majority of time is utilized for reviewing & analyzing the data received from various sensors deployed in structures. This data analysis helps in understating the structural stability and its current state with certain limitations. Considering this fact, integration with Machine Learning (ML) in SHM has attracted significant attention among researchers. This paper is principally aimed at understanding and reviewing of vast literature available in sensor-based data-driven approaches using ML. The implementation and methodology of vibration-based, vision-based monitoring, along with some of the ML algorithms used for SHM are discussed. Nevertheless, a perspective on the importance of data-driven SHM in the future is also presented. Conclusions are drawn from the review discuss the prospects and potential limitations of ML approaches in data-driven SHM applications.

As monitoring data becomes increasingly available, it is natural for structural health monitoring practitioners to turn towards data-driven models. Despite the expressive capability and flexibility of such models, their predictive performance relies on access to suitably represenative training data, which in structural health monitoring, equates to data that span the full environmental and operational envelope of the structure. Additionally, given the black-box nature of such models, there is no guarantee that predictions will adhere to fundamental physical principles. In an attempt to address these limitations, recent attention has been directed towards models that seek to combine physical insight with traditional machine learning techniques, referred to generally as physics-informed machine learning or grey-box modelling. In this paper, we seek to incorporate physical insight into a Gaussian process through the use of covariance functions that explicitly capture the evolution of dynamic systems in time. Specifically, within an autoregressive setting, we begin by deriving the covariance for the approximate response of a single degree of freedom system. Consideration is then given to how such kernels may be used when the governing parameters in the equations of motion vary over time, which is investigated here through a varying temperature, and consequently, structural stiffness. It is demonstrated that the derived grey-box models are able to outperform equivalent physics-based and data-driven models over a number of simulated case studies.

In this work, the beginning of developing a structural health monitoring (SHM) approach is presented for a representation of an aircraft composite wing spar. Lack of directly available field performance data is mitigated using a high-fidelity finite element model and a probabilistic understanding of the aerodynamic loads under different flight regimes, simulating realizations of the spar’s performance in service. Debonding damage between laminates was included in the model at different locations in the spar, with various damage sizes. Under the expectation of a fiber optic measurement system being used for data collection, the target measurements are uniaxial strain, measured in several paths throughout the spar. Given measured strain, the damage assessment problem is probabilistically formulated by defining local buckling from debonding as the observable damage, which is fundamentally characterized by load-dependent buckling eigenvalues. This FE physical model is highly computationally intensive, so machine learning was used to build a “run time” surrogate model to learn the relationships between inputs – loads and damage conditions, and outputs – strain and buckling eigenvalues. In addition, other surrogate models were created to solve the inverse problem, linking strain data to damage classification (size and location). Finally, the probabilistic frameworks are demonstrated and damage criticality assessment, which is directly related to the buckling load, is performed via Gaussian process regression.

Structural Health Monitoring (SHM) based on Lamb wave propagation is a promising technology to optimize maintenance costs, enlarge service life and improve safety of aircrafts. A large quantity of data is collected during all the life cycle of the structure under monitoring and must be analysed in real time. We propose here to use 1D-CNN to estimate the severity and the localisation of a damage with the signals measured on a composite structure monitored with piezoelectric transducers (PZT). Two architectures have been tested: one takes for input the difference of the time signals of two different states and the second takes for inputs temporal damage indexes. Those simple networks with a few layers predict with high precision the position and the severity of a damage in a composite plate. The evaluations on different cases show the robustness to simulated manufacturing uncertainties and noise. An evaluation on experimental measurement shows promising results to localise a damage on a real plate with a CNN trained with numerical data.

To address limitations presented by the unavailability of labelled data in structural health monitoring, transfer learning – in the form of domain adaptation – can facilitate leveraging information from a population of physical or numerical structures, by inferring a mapping that aligns the feature spaces. Thus, data from a similar source structure can be used to train a classifier that generalises to a target structure. Conventional unsupervised domain adaptation methods would assume that the data available from both structures belongs to the same number of damage-states or disparate environmental conditions; however, in practical scenarios, the data in the target structure may pertain to a subset of the available classes in the source structure; this is a partial domain adaptation problem. Conventional domain adaptation methods are prone to performance degradation in this scenario. To address this issue, this paper proposes a novel statistic alignment method and instance-weighting strategy. A numerical population of structures demonstrates that these methods facilitate transfer where a number of state-of-the-art domain adaptation algorithms cannot.

Health indicators are indices that act as intermediary links between raw SHM data and prognostic models. An efficient HI should satisfy prognostic requirements such as monotonicity, trendability, and prognosability in such a way that it can be effectively used as an input in a prognostic model for remaining useful life estimation. However, discovering or designing a suitable HI for composite structures is a challenging task due to the inherent complexity of the evolution of damage events in such materials. Previous research has shown that data-driven models are efficient for accomplishing this goal. Large labeled datasets, however, are normally required, and the SHM data can only be labeled, respecting prognostic requirements, after a series of nominally identical structures are tested to failure. In this paper, a semi-supervised learning approach based on implicitly imposing prognostic criteria is adopted to design a novel HI suitable. To this end, single-stiffener composite panels were subjected to compression-compression fatigue loading and monitored using acoustic emission (AE). The AE data after signal processing and feature extraction were fused using a multi-layer LSTM neural network with criteria-based hypothetical targets to generate an intelligent HI. The results confirm the performance of the proposed scenario according to the prognostic criteria.

Structural Health Monitoring deals mainly with sensorised structures providing signals related to their “health status” aiming at lower maintenance costs and weights of aircrafts.Much effort has been spent during last years on analysis techniques for evaluating metrics correlated to damages’ existence, location and extensions from signals provided by the sensors networks. Deep learning techniques can be a very powerful instrument for signals patterns reconstruction and selection but require the availability of consistent amount of both healthy and damaged structural configuration experimental data sets, with high materials and testing costs, or data reproduced by validated numerical simulations. Within this work explicit finite element simulation of impact damages on composite plates have been used to populate the data sets necessary for deep learning algorithm. The final scope is the detection of delaminations into composites plates for aerospace employ. Numerical time history signals have been collected for both the healthy and unhealthy state of the structure and transformed into RGB images. A convolutional algorithm trained on healthy and damaged signals has been used to identify anomalies in the form of delamination in the structure. This paper will present the preliminary results achieved by the authors.

We propose a real-time non-destructive evaluation technique for defect detection in composites using highly nonlinear solitary waves (HNSWs) and a deep learning algorithm based on the convolution neural network (CNN). This technique implements deep learning to identify the presence of defects and classify the defect locations in the thickness direction of composites through HNSWs with strong energy intensity and non-distortive nature. To collect HNSW datasets for training and validation of the deep learning algorithm, AS4/PEEK composite specimens with artificial delamination are fabricated and HNSW datasets are generated from the experimental setup of a granular crystal sensor. Testing pretrained CNN based algorithms verifies the performance of detecting and classifying defects by location in composite plates.

In the context of Guided Wave-based Structural Health Monitoring (GW-SHM), ultrasonic elastic waves are used to detect damages in structures by comparing the acquired signals with those from a defect-free structure. However, the high sensitivity of GWs to environmental and operational conditions limits the validity of such references. Notably, variabilities between multiple specimens are often significant from the GWs perspective. These variabilities are particularly important in composites and are due to sensor positioning, sensor coupling and material variability. This communication presents a baseline-free approach using physics-enhanced Machine Learning (ML) for enhanced robustness. To ensure the coverage of these variabilities the approach is validated on multiple Carbon-fiber-reinforced polymers (CFRP) panels. The methodology relies on feature extraction from raw GW signals and training classification algorithms (e.g., kernel machines, neural networks). To make the classifier learn inter-specimen variabilities, an experimental database of 45 impacted composite panels is used. Half of them are used to provide pristine data, and the rest to provide damaged data so that the same sample is either in the training or test set, but never in both. Good classification performance is obtained, demonstrating that the classifier has successfully learnt to recognize defect signatures despite the variability linked to the multiple specimens and instrumentations.

Acoustic emission testing remains a popular technique in the SHM community due to its effectiveness in detecting and localising small occurrences of damage in a structure. One of the characteristics of acoustic emissions is that all structures will produce some level of emission even if only benign flaws are present. As such, it is often necessary to collect a baseline reference set of data against which newly collected measurements can be compared to establish the state of the structure. In this work, a methodology is proposed which eliminates the need for this “training phase” through use of a statistical model which learns and adapts online. The chosen technique is a Bayesian online changepoint detection method where bursts of acoustic emission are modelled as a Poisson point process. In this way the inherent stochasticity in the number of “hits” emitted in a given window is learnt and modelled online, then significant changes in the properties of the generating stochastic model are used to provide sensitivity to damage. This approach has a number of benefits which are demonstrated on an experimental dataset from a bearing test rig. The main benefit is removing the need to collect extensive data before the SHM system can become operational. The proposed solution also allows prior engineering knowledge to be exploited, for instance by specifying priors related to the expected number of hits in a given window, or through use of a Hazard Function which encodes prior belief about the expected time before damage will occur. Finally, the method is shown to characterise the full probability distribution over possible run lengths. This information provides not only an indication of if a significant change in the behaviour of the system has occurred, but also automatically quantifies a degree of confidence in that change.

Fatigue is a common cause of the failure of structures: previous work by the authors has shown the usefulness of using Gaussian process regression to develop a probabilistic assessment of fatigue damage accumulation. By propagating uncertainty from a predictive model for structural response (strain) under unknown loading, a more robust assessment of the damage state of a structure is enabled. Although these black-box models have previously shown good results for quasi-static problems, dynamic behaviour is difficult to predict in this way. Explored here is a promising and novel means of accounting for this, by integrating physical knowledge specifically through the GP kernel. The impact of this on accuracy of fatigue damage prediction is shown to be significant and the damage variance from a probabilistic perspective is reduced substantially.

Hybrid approaches for training machine learning algorithms to identify damage in bridges rely on the use of both monitoring and numerical data. While monitoring data account for normal operational conditions of the undamaged structure, numerical data are often confined to scenarios that seldom occur in the lifespan of the bridge, like extreme temperature events or damage, although previous research of the authors showed it can also be used to augment the data acquired under regular service. This paper presents a hybrid approach for damage identification and applies it to Z-24 Bridge. To enable the classification of damage, supervised learning algorithms are employed. Unlike unsupervised learning, which relies on unassigned data and is suited for novelty detection, supervised learning uses labeled data corresponding to undamaged and damaged scenarios of the structure, enabling the transition from damage detection and localization to damage type and severity. A hybrid database is constructed using monitoring and numerical data corresponding to undamaged scenarios and numerical data corresponding to damage scenarios. The damage scenarios comprise various degrees of settlement of a bridge pier and a landslide near the same pier. Several common supervised learning algorithms are trained with the hybrid data and a comparison of the results is provided.

Damage diagnosis for offshore wind turbine foundations is a topic that remains open in the scientific community due to the importance of increasing safety and ensuring functionality. To deal with the challenge of online and in-service structural health monitoring (SHM) for wind turbines, approaches based on the vibration-response of the structure and captured by sensors such as the accelerometers need to be considered. This work presents a novel methodology to improve the structural damage classification of wind turbine foundations. This methodology consists of several stages. First, the data acquisition to collect and organize the information from the sensors attached to the structure, following the use of a mean-centered group scaling (MCGS) procedure to normalize the raw data and eliminate the difference between the magnitudes of the sensors. Next, a data unfolding to allow the multivariable analysis is performed. Then a linear feature extraction stage is applied to reduce the high dimensionality of the signals. Subsequently, the new feature array serves as the input to a supervised machine learning algorithm which allow to perform the classification task. A five-fold cross-validation procedure is used to obtain the goodness of classification. Several classification performance measures are calculated considering an imbalanced data set obtained with experimental data of a small-scale wind turbine foundation structure to validate the results of the proposed methodology.

The rotor-bearing-squeeze film damper coupling (RBSC) system is the core part of aero-engines, determining operation safety and reliability. The dynamic behavior of the RBSC system is easily affected by the bearing outer race localized defect, which is difficult to identify accurately. Inappropriate dynamic analysis of system behavior will further affect monitoring and diagnosis of the system. This paper presents a mathematical model of the RBSC system with bearing outer race localized defect to study the dynamic behavior of the system. The model develops a smooth spalling path in the bearing outer race to simulate the successive passage of the balls. The model also considers the effect of centrifugal load on the bearing force. The impact of the bearing outer race localized defect on the RBSC system during the acceleration process is analyzed. The numerical results show the bearing outer race localized defect will aggravate the nonlinear behavior of the RBSC system and is prone to chaotic motion. Meanwhile, affected by the bearing outer race localized defect, the vibration of the RBSC system will increase in the speed range of relatively low and relatively high. Especially in the high-speed range, the impact of a bearing outer race localized defect on system vibration covers more speeds. This study provides a theoretical basis for condition monitoring and fault diagnosis of the RBSC system.

Wave propagation in structures is generally computed by numerical methods such as finite element method, spectral element method, etc. In these numerical methods, spatio-temporal discretization of the partial differential equations is performed using a fine mesh which leads to high computation cost but precise results. A trade-off between accuracy and computational cost can be achieved by adopting deep learning-based approaches. This research demonstrates an alternative deep learning-based approach for predictive modeling of wave propagation signals within damaged structural elements. Our goal is to evaluate the wave propagation spatio-temporal solution matrix for a given crack depth and crack location within the structural element. In this framework, deep-learning-based surrogate modeling is proposed by utilizing a deep convolutional autoencoder (DCAE) to learn the wavefield representation and project it to a compressed domain called latent space. This latent space works as labels for a feed-forward neural network (FFNN) followed by DCAE. This process eliminates the need to solve the system’s governing equations each time, leading to significant savings in computational costs and making the method excellent for issues that require repeated model computations. In DCAE architecture, we integrated the squeeze-and-excitation (SE) block which works as a channel-wise attention mechanism and enhanced the performance of the model. The DCAE with SE block achieved the very good reconstruction accuracy. This deep learning-based wave propagation predictive model can be a valuable resource for generating data for a given crack depth and location, which can be used for inverse formulations and various structural health monitoring (SHM) application.

In this paper, we present a deep learning technique for image segmentation known as the global convolutional network, which we employ for delamination detection and localisation in composite materials. The model was trained and validated on our previously generated dataset that resembles full wavefield measurements acquired by a scanning laser Doppler vibrometer. Additionally, the model was verified on experimentally acquired data with a Teflon insert representing delamination, showing that the developed model can be used for delamination size estimation. The achieved accuracy in the current implemented model surpasses the accuracy of previous models with an improvement of up to $$22\%$$ 22 % for delamination identification.

To reduce the operation and maintenance cost for wind farms, turbine operators are actively developing strategies to, among others, reduce service cost, maximise power production, and prolong lifetime of components and super-structure. All of these tasks require a wind farm model that can accurately predict turbine behaviours in response to the changing environment. Recent studies focus on developing data-based methods for predictive maintenance purposes. This paper proposes a data-based model that aims to capture the spatial and temporal wind variations across a wind farm, as a means to predict the interactions between operating turbines and the environment, which can be useful for wind farm performance monitoring. The proposed method is a Gaussian process-based spatial autoregressive model, which reflects our physical understanding of the wake effect while taking the advantage of a stochastic data-driven learner. In the case study of a simulated wind farm, the proposed model (named here a GP-SPARX model) provides the best predictive accuracy in comparison to two other spatial autoregressive regression models, showing its capability of capturing nonlinear correlations and its potential as a low-cost wake field predictor given inputs from weather station measurements.

Composite materials are prone to various kinds of defects in their service life, among which delamination is a very hazardous type of damage. The traditional visual inspection techniques often fail to detect delamination in composite structures. Guided Lamb waves are increasingly being applied for the identification of delamination in these structures. Scanning laser Doppler vibrometry can measure the full wavefield of guided Lamb waves, such full wavefield contains rich information about defects. In this research work, a novel deep learning-based semantic segmentation technique is applied for delamination identification on full wavefield data. A big dataset of full wavefield images resulting from the interaction with delamination of random shape, size, and location was utilised and fed into the proposed deep learning model. The main motive of this research work is to investigate the applicability of deep learning-based approach for delamination identification in composite structures by using only the animations of guided Lamb waves. It is verified that the performance of the proposed deep learning model is good. Moreover, it enables better automation of identification of delamination, which can further produce damage maps without the intervention of the user. Furthermore, the developed deep learning model also indicates the capability of generalising well to the experimental data.

Zaruma, a well-known mining district in the province of El Oro in Ecuador, suffers from subsidence phenomena related to illegal underground mining. Mining in this area goes back to the pre-Hispanic time; recorded history shows the first settlements of the Cañari culture and thru the Inca culture, and passing from the Spanish, American, and Canadian presence in the area up to modern times. Mining in this district has been in a constant transition from legal to illegal mining, technical and non-technical mining, prevailing activities in the present are illegal and non-technical over what is declared as the non-exploitation area under the city of Zaruma. This brings a series of problems among this land subsidence affecting the city cited as a cultural heritage cite; subsidence and collapses have accrued sporadically in the past, but in recent years (since 2017) this phenomenon has become a relatively consistent manner. In fact, December 15th, 2021, is the most recent sinkhole event resulting in the evacuation of over 300 people and the banning of over ten city blocks.For this reason, it is essential to find a quick and economical technique that can generate information about the spatial and temporal development of uncontrolled underground activities to improve risk management. In this work, the Differential Interferometry Synthetic Aperture Radar (DInSAR) technique, implemented in the SUBSIDENCE software, has been used to study terrain deformation related to illegal artisanal mining in Ecuador [1]. This work presents an up-to-date study of the monitoring and detection of subsidence phenomena in the city of Zaruma, as part of a collaboration project between local and international Universities allowing to detection and monitor surface deformations using DInSAR techniques as a tool applied to monitoring mining-related subsidence phenomena.

Great part of the existing constructions has reached or overcome its service life. A structural assessment is not always necessary, nevertheless, it is important to monitor these structures, both for their maintenance and preservation of human lives purposes. Remote sensing techniques have proved to be useful in supporting the management process and the safety evaluation, by reducing the impact of disturbances on the construction functionality. Also, the reconstruction of the geological and stratigraphic setting of the soil beneath the construction is fundamental to evaluate any correlation between the ground movements and their causes. In this work, Differential Interferometric Synthetic Aperture Radar data are used to detect the ground movements experienced by the construction. These data are then integrated in GIS environment with the information about the engineering-geology of the area. The geometry of the structure, based on historical documents, and the subsoil stratigraphy have been modelled in 3D. The procedure is applied to the residential area of Valco San Paolo in Rome. Once the 3D model is constructed, a preliminary structural health monitoring can be performed, providing insights on critical displacement trends. The maximum deformation parameters are compared with limits derived from literature to identify the damage thresholds.

In Structural Health Monitoring for a real understanding of the changes taking place and their effects on the structural integrity of the built environment, sometimes it is necessary to move to long observation times. This follows the axioms of SHM, which identify a certain relationship between the frequency and time of observation and the extent (and therefore severity) of ongoing damage. For this reason, in the present paper, interferometric displacement satellite data acquired for a decade on the territory of the city of Rome (Italy) are investigated and correlated to natural phenomena. The paper critically analyzes the possibility of a relationship between these phenomena and satellite data in order to bring out a common pattern. The study of natural and anthropogenic phenomena in the same frequency bandwidth as the interferometric satellite data would therefore be useful for recognizing potential triggering causes of higher frequency phenomena, which could appear as sudden and unstable phenomena if observed with shorter times. In the paper, the authors first make a comparison between natural phenomena and satellite data on a territorial scale and then focus on a series of isolated case studies (single structures and infrastructures).

The management process and safety evaluation of existing buildings, which have often overcome their service life, can be properly supported by non-invasive structural monitoring techniques, among which multi-temporal Differential Synthetic Aperture Radar Interferometry (MT-DInSAR) techniques can be included. The paper follows previous works of some of the authors on this topic, giving an insight on the proper use, processing and interpretation of satellite radar interferometry data for the structural analysis, through an application to the “Vittorino da Feltre” school building located in the city centre of Rome (Italy), within the framework of a retrofitting project. After performing a correct positioning of the reflecting targets on the 3D geometry of the structure under investigation, displacement time series, mean deformation velocity values and continuous maps of the vertical and East-West velocity components are presented, by exploiting the COSMO-SkyMed SAR data collected during the 2011–2019 time interval. Finally, a comparison of the obtained results with the on-site detected cracking pattern can be useful to better understand the ongoing phenomena, for a proper damage assessment.

Structural Health Monitoring (SHM) is a field of increasing interest and worthy of new approaches and innovative applications. As well known, Italy boasts a unique cultural-historical heritage of monuments and archeological sites that need to be managed, particularly in a multi-hazard prone area like our country. A really appealing technique is represented by the advanced multi-temporal differential interferometry synthetic aperture radar (MT-DInSAR), which has been developed and applied in different fields in the last twenty years. The exploitation of such techniques in the monitoring and structural assessment of cultural heritage is still an open issue even if the first applications available in literature show promising potentialities. In this paper the general framework for structural monitoring and assessment previously proposed by the authors, is applied to the complex building San Michele in Rome (Italy). In particular, COSMO-SkyMed (CSK) ascending and descending datasets are collected and processed applying the Small Baseline Subset (SBAS) method obtaining deformation time series and mean velocity maps of the persistent scatterers located in the investigated area, for both geometry acquisitions. Finally, different techniques useful for assessing the structural behavior and monitoring of constructions are applied and critically discussed.

Following the concept that structures move because something causes their movement, in the paper, the authors analyze interferometric satellite Line-of-Sight displacement acquired over the territory of Rome (Italy) during the period 2011–2019 from a spectral point of view. To obtain a comprehensive understanding of the behavior of civil structures in this unusual frequency bandwidth (~2–200 nHz), an in-depth analysis of the possible sources of perturbation is called into question.To reach this goal, the difference in the vibration characteristics of measured Line-of-Sight displacements and spectral entropy is studied. The main environmental effects that could affect the built environment (e.g., temperature, hydraulic balance, etc.) are also analyzed in terms of the power spectrum and spectral entropy. The main frequencies associated with both environmental data and remote displacements are highlighted.A more in-depth knowledge of the inputs of a system would allow to perform more significant virtual experiments or numerical predictions, also contemplating new sources of excitation for civil structures at nanohertz.

The application of Optimal Sensor Placement (OSP) algorithms has been advancing in Structural Health Monitoring (SHM). Among many approaches, this study focuses on a Bayesian OSP algorithm optimized using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) – a blackbox algorithm for continuous problems. The increase in the number of CMA-ES variants warrants the non-trivial selection problem. Nonetheless, it is difficult to say if such a step has been well-conducted in the investigated SHM literature. Hence, this study aimed to improve the existing OSP framework by adding an offline algorithm selection methodology and to demonstrate the framework’s application to buildings. Application to the ASCE benchmark structure showed that the discrete implementation of CMA-ES outperforms a discrete reformulation that used binomial sampling. OSP results were also found to agree with existing literature. Furthermore, case studies quantify that for an undamaged structure, sensor placement that is well-distributed according to the location of the columns and stiffeners gives high-quality OMA results. For damaged structures, sensors should be placed near the damaged region, and failure to do so showed a reduction of MAC value by 0.1. Lastly, with insight from industry research of the Philippine SHM landscape, these conclusions were found to be potential recommendations to improve the current developing practice compared to the international state-of-the-art.

Structural Health Monitoring methods are in general based on material testing during an operation of structures under monitoring. We propose a new method based on use of Small Punch Test, which does not require repair of the structure after sampling unlike classical testing methods. The way of material sampling for SPT makes it a non-destructive testing method thus allowing material condition assessment at any stage of machine life without the need to exclude it from usage and cause production downtime. In this work we investigate the influence of 250000h of work in extreme conditions on 10H2M material with the use of Small Punch Test methods. The material has been tested in both states: heavily fatigued and unused to prove SPT usability for the industry. To cross compare the results a Uniaxial Tensile Test has been conducted as well on both versions of the material. Small Punch testing methods included the assessment of Yield Stress, Ultimate Strength, and Ductile-Brittle Transition Temperature. A statistical approach has been used for results analysis to determine SPT accuracy in comparison to the standard testing methods.

Structural Health Monitoring (SHM) is gaining increasing attention in Italy and worldwide due to structural obsolescence and sudden collapses occurring from time to time due to insufficient maintenance or extreme events. On the other hand, the technological progress in the SHM field is making it particularly attractive as a complement to visual inspections and in-situ surveys aimed at assessing the structural safety. Accordingly, several guidelines have been developed with the aim to provide useful recommendations to technician for the design of SHM systems. Nevertheless, because of very case-specific design, so far, a general qualification procedure aimed at assessing the performance of a SHM system is still missing. On the contrary, construction products already share a thorough and well-established harmonized standardization framework since many years, and this resulted in a reliable control of performance. In this study, a preliminary qualification approach for SHM systems is proposed. The qualification scheme is scenario dependent and allows to check the effectiveness of a given SHM system defined in terms of hardware as well as software components. In order to validate the approach, different SHM systems are hypothesized and checked for possible qualification with respect to different scenarios, obtaining encouraging results. The proposed approach, therefore, represents a promising attempt towards a more exhaustive and comprehensive qualification framework for civil SHM applications.

The growing demand towards life cycle sustainability has created a tremendous interest in non-destructive evaluation (NDE) to minimize manufacturing defects and waste, and to improve maintenance and extend service life. Applications of Magnetic Sensors (MSs) in NDE of civil engineering structures have become of great interest in recent years due to their non-contact data collection, and their high sensitivity under the influence of external stimuli such as strain, temperature, and humidity, to detect damage and deficiencies. There have been several advancements in MSs over the years for strain evaluation, corrosion monitoring, etc. based on the magnetic property changes. However, these MSs are at their nascent stages of development, and thus, there are several challenges that exist. This paper summarizes the recent advancements in MSs and their applications in civil engineering. Principle functions of different MSs are discussed, and their comparative characteristics are presented. The research challenges are highlighted and the roadmap towards high technology readiness level is discussed.

The paper presents the realization and testing of concrete modular prototypes reinforced with textile - steel combination and sensorized with distributed Fiber Optic Sensors (FOS). Proper procedures for fastening and pre-tensioning the sensor to the textile, as well as for protecting the sensor during and after concrete casting, were implemented. Optical Time Domain Reflectometry (OTDR) tests confirmed the healthy status of the optical fibers and related connectors after 14 months of exposure in tunnel environment. One of the prototypes was tested in flexure, investigating the effectiveness of strain monitoring by means of the embedded distributed FOS, interrogated by Brillouin technique with enhanced spatial resolution. The paper discusses the prototypes preparation and the test results, leading to the successful demonstration of the concept of sensorized modular structure, provided with built-in monitoring capability and easily connectable on the field by means of the optical connection of each module to the adjacent one.

Bridges are critical infrastructures subjected to cyclic loading and require fatigue monitoring to prevent high maintenance costs due to fatigue failure. This paper, part of OWI-lab’s research activities within the SafeLife-Infrabel project, presents a fatigue survey on four months of strain measurements of a steel railway bridge in Belgium, comprising 98 Fiber Bragg Gratings (FBGs). This study aims to develop a data-driven case/event detection scheme including measurements and operational data (e.g., train type and passage time). The first objective is to develop a Python package to separate the events by automatically selecting the train passage events from the strain time series to analyze them and also reduce the dataset size. Over the studied period, a total of 5000 events were detected. Then, the available operational data is complemented with properties estimated from the strain measurements, including the axle number, speed, and direction. Finally, the relation between the fatigue damage and different train features is studied by calculating the attributed fatigue damage using the Rainflow cycle counting method and the Palmgren-Miner rule. A notable damage difference existed between freight and passenger trains. In addition, the damage difference between loaded and empty freight trains was completely distinguishable, while the effect of occupancy was not very visible in passenger trains. Axle number had the highest impact among the passenger trains and had linear relation with damage. Also, the difference in fatigue damage between various types of passenger trains was less distinctive. Finally, the speed and direction affected the damage very slightly.

Elastomeric bearings are ubiquitous in railway bridges as they provide flexibility that prevents stress concentrations near the supports due to temperature-related material contraction and expansion. Their inherent flexibility affects the overall stiffness and the vibration frequencies of the bridges. Furthermore, the stiffness of the bearings can vary significantly during the lifetime of the bridge due to several factors such as material deterioration and aging and can directly impact the behavior of the bridge. This article investigates the impact of the variations in the stiffness of elastic bearings on the train-induced vibrations on single-span railway bridges. A series of parametric analysis for different train speeds was conducted. The analysis were repeated for a wide-range of bearing stiffness that represents the variation of this parameter during the lifetime of the bridge. The results indicate that, the variations in the bearing stiffness due to aging and deterioration have significant impact on the train-induced vibrations.

Masonry arch bridges are numerous across European transportation networks. Many are ageing structures, with service lives of 100–150 years to date, and exhibit historic damage and repairs, leading to uncertainty regarding structural behaviour. For skewed bridges particularly, this can be complicated and three-dimensional, and detailed experimental data describing behaviour are rare.In 2018–2019, the authors deployed Fibre Bragg Grating (FBG) strain monitoring at a recently repaired, skewed masonry rail bridge in the UK. Following an on-site trial, the FBG monitoring system was substantially upgraded in 2020 to enable long-term, autonomous, remote sensing. This new system is introduced, including processes to automate data classification based on the date and time of measurements, and train class/operator, direction, and speed.This system has recorded the bridge responses to thousands of trains. Data analysis is presented, focusing particularly on seasonal and long-term variation of behaviour. Findings include the impact of ambient temperature; an inverse relationship is observed. Decreasing temperature causes thermal contraction of the masonry, allowing cracks to open and increasing the potential for bridge movements. After decoupling such effects, residual long-term changes may correspond to damage. Therefore, this system can provide valuable asset management information on the early onset of bridge deterioration.

Damping is an inherent structural property that plays an important role in the overall behavior of dynamical systems, and it can also represent a potential indicator of their health. However, its estimation is often affected by significant uncertainties, which are attributable to measurement noise, test conditions (e.g., environment, level and type of applied loads), and identification technique. In this contribution, a novel procedure is developed for modal damping ratio identification from free vibration tests, which exploits a tuned version of the Variational Mode Decomposition (VMD) technique for signal decomposition. Once the time history response of each mode is extracted, relations among areas forming from the free-decay time-domain response are used to estimate the modal damping ratio of each mode. This approach mitigates the detrimental effect of noise contamination likely occurring in real signals. Two applications are presented to illustrate the efficiency of the proposed identification strategy. The first example deals with the analysis of a theoretical signal (with pre-defined frequencies and damping ratios) for validation purposes. A real signal related to the free vibration response of a prestressed concrete bridge deck is finally examined. Based on the obtained results, the proposed identification strategy proved to be efficient and suitable for automatic implementations, thereby resulting particularly attractive for practical applications.

With more than 600,000 highway bridges, 46.4% of which rated as fair and 7.6% rated in poor condition, United States is one of those countries in which the installation of reliable bridge health monitoring systems is strategically necessary to minimize and optimize repair and rehabilitation costs and to prevent failures. In this review paper, a synthesis of the scientific literature relative to the SHM systems installed in some U.S. bridges over the last 20 years is presented. This review aims to offer interested readers a holistic perspective of recent and current state-of-the-art in bridge health monitoring systems and to extract a “general paradigm” that is common to many real structures. The review, conducted through a comprehensive search of peer-reviewed documents available in the scientific literature, discusses several bridges in terms of the instrumentation used, scope of the monitoring, and main outcomes.

Distributed fiber optic sensing (DFOS) is commonly used to monitor large structures like bridges or tunnels. However, DFOS can also be beneficial for smaller structures, even single family houses. This articles discusses a real world example where sensing cables were embedded in a two story house. By applying different fiber optic sensing techniques, various information can be gathered using the same DFOS cables.Distributed temperature sensing (DTS) enabled the assessment of the heat insulation of the ground plate of the house. Potential weak points can be clearly identified and repair works can be made at the exact location of the problem. Distributed strain sensing (DSS) was used to monitor the swelling and shrinkage of the concrete. In addition, numerical simulations of the ceiling bending were verified with DSS. Distributed acoustic sensing (DAS) was applied for intruder detection. It is shown that walking of people in the house could be detected and that their location was pinpointed to individual rooms.In this article, the installation, the data acquisition and analysis of the different sensing techniques are discussed. An outlook is given on how the presented approaches can be scaled from the monitoring of a single house to an apartment building.

Visual inspection and manual documentation of masonry is a time consuming and subjective process. This paper aims to improve automation in documentation and assessment of ageing masonry infrastructure through image-based techniques and machine learning. A large dataset of annotated images has been developed to train a deep learning model. Different convolutional neural networks were investigated to identify the most suitable for the task. The results presented are combined with previous work to generate high quality geometrical and numerical models of masonry infrastructure. This implementation of deep learning, for segmentation and localisation of bricks in masonry, highlights the potential of recent technologies for the automation of structural inspection, documentation, and analysis of cultural heritage.

In this paper the potentiality of an acoustic camera coupled with a machine learning algorithm to detect possible anomalies of an operating gear box is investigated. First, an experimental campaign was performed for different operating conditions (velocity, amplitude, frequency). During these phases the sound images were collected with the acoustic camera. This is followed by the pre-processing phase in which the acoustic images are prepared to train the network. The next step concerns the creation of a Convolution Neural Network (CNN) suitable for the classification of sound images. The last one involves training and testing of the network created.The analysis of the training plot and the confusion matrix show promising results. Most of the analyzed images are classified correctly with an overall accuracy of the model of 95%, despite the simplicity of the network created. Observing the excellent obtained results, this technique promises to be suitable for non-intrusive monitoring, allowing companies to reduce maintenance costs.The strength of this procedure is that, although the measurements are made in a noisy environment and not in an anechoic chamber, the Convolutional Neural Network is able to classify the images very well.

Electrical actuation concept fits within the worldwide strategy of more electric aircraft. However, the integration of linear electromechanical actuators is challenging in safety critical systems due to jamming of the driven load. That fault is critical in many applications such as primary flight controls or landing gears extension and steering. This paper reviews solutions currently available for aerospace application, limits for their effective deployment and possible solutions to achieve a condition-based maintenance to reduce costs still guaranteeing safety. Typical conversion mechanisms adopted so far are described along with the failure modes of these systems, contrasting their many advantages. Particular attention is given to the jamming and possible strategies to avoid it. In particular, any structural alteration which can induce jamming can be monitored as prescribed by predictive maintenance approach in order to detect upcoming faults in advance. Among different possibilities available to timely assess latent fault inducing jamming, the analysis of electrical current and/or mechanical vibrations signals by deterministic or statistic approaches are the most promising. In addition they show strong potentialities in implementing data fusion and learning algorithms to identify different EMA operational conditions.

Pipelines are often used in chemical plants, oil refineries and power plants. For economic reasons, there is a requirement to monitor a long section of the pipelines. In this work, we present a monitoring approach for metallic pipelines using actively induced microwaves in the 1–5 GHz frequency range. This work presents numerical as well as experimental damage detection results. This also includes the design, fabrication and characterization of a broadband antenna structure. In addition, an experimental setup based on a stainless steel pipeline will be presented. Data acquisition is performed by a vector network analyzer. This paper presents initial experimental results for monitoring of metallic pipelines using a reproducible damage model.

This work aims at a structural health monitoring (SHM) system using radar sensors at 60 GHz embedded in a wind turbine blade. For a reliable operation of the system, the influences of changes in the ambient climate on the embedded sensors has to be taken into account and shall thus be studied and analyzed. Three sensors are embedded for experiments and are exposed to a defined environmental state in a climate chamber. The signals for different target states are measured, processed and discussed. An evaluation of the dependencies shows how critical these influences are on radar signals and baseline-based SHM methods working with them. These results motivate the development and classification of compensation methods. The considered methods are a baseline selection method and a polynomial fitting approach. Both methods are exemplarily applied to the measured data. Finally the sensors are used to measure ice formation. Conclusions on the detectability of ice are made, as well as on its influence to other measurements.

Additively manufactured (AM) parts generally have a higher surface roughness than parts manufactured using conventional methods. In most applications, a smooth surface finishing is preferred since rough surfaces are prone to corrosion attacks and fatigue crack incubation. Therefore, surface finishing is necessary to reduce the surface roughness of AM parts before they are used. The roughness of outer surfaces can be measured directly, using either contact stylus profilometer or non-contact optical microscope. However, there are still challenges to quantify the roughness of the internal surfaces using these conventional techniques of surface metrology. In this paper, we present a method to measure the roughness of internal channels by analyzing ultrasonic signals from the backwall reflection. A frequency-domain technique based on phase-screen approximation is used to reconstruct the root mean square ( $$R_{q}$$ R q ) value from the ultrasonic signal scattered from the rough surfaces. Finite element simulations are developed to demonstrate the method, showing excellent accuracy when the $$R_{q}$$ R q value is within 1/15 of the wavelength of the incident ultrasound. The method is then applied to monitor the surface roughness of the internal channels during the Abrasive Flow Machining (AFM) process. The reconstructed roughness value shows a clear and steady downward trend during the polishing process, quantitatively indicating the polishing rate. This demonstrates that such ultrasonic method can be used as a tool to provide feedback controls in the polishing process.

In the field of structural health monitoring (SHM), the use of ultrasonic transducers has been increasingly prevalent for thickness gauging and corrosion monitoring. Since temperature affects ultrasonic velocity, temperature fluctuations and non-uniform temperature distributions in components are a major source of uncertainty for ultrasonic measurements. Recently, ultrasound-based temperature sensing methods have gained increasing attention because of their ability to measure subsurface temperature distributions in solid media. While conventional ultrasonic techniques can either measure the thickness of a component at constant temperature or the (subsurface) temperature of the component at constant thickness, measurement accuracy and precision can be greatly compromised if both factors vary simultaneously. In this work, we explore a dual wave approach to overcome this limitation of the conventional methods. Results of simulation and experimental studies show that co-located shear and longitudinal pulse echo measurements can be used to determine both the thickness and through-thickness temperature gradient of a steel plate while the structure is undergoing thickness loss at rapidly changing temperature.

The reliability of MUT is arguably sub-par. There has been very limited work on the reliability of MUT of corrosion as most of the research focused on MUT of weldments. The current body of literature showed a host of human factors that affected MUT, and a survey carried out on Non-destructive testing (NDT) personnel working in industry highlighted that two areas of concern were the task and the inspector. It is proposed that the use of a magnetic probe coupler (MPC) will aid the task and the inspector in carrying out a reliable MUT inspection. It will do this by ensuring that probe coupling is maintained to the Carbon Steel surface when working in a variety of positions while improving the tactility of scanning a component. It will greatly aid the monitoring of known areas of wall loss by ensuring that the probe is coupled consistently and accurately in the same position. The areas of wall loss can be easily marked when the technician is not manually trying to hold the probe to the test piece. The research will look to demonstrate the effectiveness of the MPC utilising it in corrosion inspections trials to validate the MPC.

Icing causes degradation of the flight performance, especially for Unmanned Aerial Vehicles (UAVs) due to lower Reynolds flight number. The necessity of a low power de-icing system is crucial for electric powered UAVs with limited energy reserves. In order to improve the efficiency of an ultrasonic de-icing system, in this work an adaptive impedance matching network is proposed. Such a system autonomously evaluates the electro-mechanical impedance of the structure to be protected by ultrasound and on which actuators are distributed. Afterwards, it finds the best configuration to minimize the reflected power when the transducers are excited at the selected working frequency. Preliminary results for a single actuator are discussed to demonstrate the system effectiveness.

The use of ultrasonic waves in the context of SHM offers methods to analyze materials and systems. Both Acoustic Emission-based approaches (passive, active) are limited by the propagation characteristics of ultrasonic waves, especially in inhomogeneous materials like carbon fiber reinforced polymers (CFRP). The use of the piezoelectric and inverse piezoelectric effect is a very accurate method of sensing and exciting ultrasonic waves. However, the transducers resonance characteristics affect the waveforms. For illustration in this contribution different excitation signals are experimentally compared in frequency domain by fast Fourier transform (FFT) and in time-frequency domain by continuous wavelet transform (CWT). Then transducers effects along the propagation path of the wave are investigated. Frequency and fiber direction dependent damping factors of ultrasonic waves in CFRP as well as the influence of the transducers are determined. The distance between sensors in a sensor network is limited by attenuation, so fiber direction must be considered. Finally, by analyzing the frequency response of the transducer, a filtering method is developed to compensate for the resonance characteristics of the transducers. Finally, a more accurate estimate of the energy released and therefore a more accurate estimate of the severity of damages/failures is proposed.

The structural component evaluated in this research is a simplified representation of composite skin-stiffener joint found in a helicopter tail boom, which was monitored during quasi-static loading conditions using distributed fibre optic sensors, along with resistance wire strain gauges and flash thermography. The skin and the stiffeners were fabricated out of unidirectional carbon fibre prepreg (CYCOM G40-800/5276-1) from Cytec and were bonded with AF163-2K structural film adhesive from 3M. The path chosen for the fibre installation was designed to provide strain response for comparison with the strain gauges, as well as to witness any change in the strain due to disbonding between the stiffener and the skin. The component was quasi-statically loaded until the skin-stiffener joint fully disbonded. The fibre optic sensor was able to detect and track the damage growth during quasi-static testing by monitoring variations in the strain distribution at the skin and stiffener interface. Results from the fibre optic sensor corroborated well with the inspection carried out using flash thermography.

Computer Vision in general and Phase-Based Motion Magnification (PBMM) in particular have been proved in recent years to be feasible and practical for the structural integrity assessment of simple laboratory models under controlled environment conditions. This technique has been partially tested for outdoor applications under natural illumination. In this paper, the potentialities of these enabling technologies for on-site surveying are highlighted, especially focussing on typical applications for Civil Engineering. These include two case studies, both of relevant interest due to their structural peculiarities: a 400+ meters high skyscraper and a concrete highway bridge. Advantages and limitations for both these applications are highlighted. The comments are then extended from these particular cases to some broader considerations valid for other structures and infrastructures.

In order to ensure safe shipping traffic with the overaged waterway infrastructure, structure health monitoring for massive hydraulic structures is becoming increasingly important. The necessity of a monitoring system usually results from one of two reasons: Either the calculated bearing capacity can be verified, but there is damage (e.g. cracks) that endangers the bearing capacity, or the structure has shown normal behavior since construction but cannot be verified by calculation. As a result of the significantly lower hazard potential of the waterway infrastructure compared to dams, automatic measuring systems are usually not installed but individually conceived if necessary. In order to optimize this process, a modular system is being developed. The paper focuses on the requirements and characteristics of massive hydraulic structures with regard to the monitoring system. These are in particular the large dimensions and individuality of the structures, the comparatively small deformations without significant vibration behavior, the long measurement periods, as well as the rough environmental conditions (humidity, temperature variations, direct solar radiation etc.). As an example, the long-term behavior of the measurements of a monitoring system in a lock is analyzed and a test rig is conceived for the investigation of the suitability and long-term stability of different inclinometer.

There is a continuously increasing demand for non-invasive damage detection methods in aerospace certification testing. This demand is driven by increasing cost effectiveness of these tests by reducing down-time for inspection. This paper aims to evaluate the usefulness of using an audible acoustics camera to detect and locate damage within a relatively large structure under bending load. An acoustic camera is a microphone array of known dimensions, which uses a ‘delay-and-sum’ beamforming numerical model, can detect sources of sound. Under the assumption that damage produces sound, this tool should aid in detecting damage. This method of damage detection was validated against strain gauge, LVDT data, and video recording showing the damage propagation. This paper finds acoustic cameras to have a potential to become a stand-alone detection tool. The authors suggest potential developments to the existing system which could potentially produce a more comprehensive method.

The use of highly nonlinear solitary waves (HNSW) to inspect structural properties is a new methodology with great promise. One application of HNSW is to measure intraocular pressure (IOP). This is accomplished with HNSW transducers designed using a chain of particles stacked in Hertzian contact. In consideration of the anatomy of the human eye, the particles and transducer are scaled down in size compared to those used in previous designs. With the end goal of a designing a small, mobile transducer, all electronic components need to be compact, and battery powered. The transducer design also needs to trigger and measure the HNSWs. To accomplish this, a particle chain containing ferromagnetic particles surrounded by an applied magnetic field is utilized. To start the wave, a striker particle is lifted and dropped by a solenoid. Controlling the solenoid is the first step in ensuring consistent HNSW waves which requires a driver to control the mechanism. Once triggered, HNSW travel through the particles causing a distortion in the magnetic field that is measured by an inductor coil. This paper will present the design of a new compact HNSW transducer and an experimental evaluation of the solenoid drivers and their current consumption.

The recording and processing of acoustic emissions can be used to identify and localise damage mechanisms occurring in engineering structures. In plate-like structures, acoustic emissions propagate through the structure as guided waves. With a measurement location away from the source location, dispersion effects in the guided wave distort the acoustic emission signal. The distortion of the original signal hampers identification of damage mechanisms.This research describes and assesses a method to reconstruct the original acoustic emission signal using dispersion compensation. Simulations and experiments are performed involving thick glass-fibre reinforced plastic laminates. The signal reconstruction on the simulated data gives a reasonable representation of the simulated signal at the location of interest. In the experimental case, similarity slightly degrades. Deviation in arrival time between original measurement and reconstruction is attributed to a possible discrepancy in material properties in reality versus the properties used in the reconstruction.

During the life of carbon fiber reinforced plastic (CFRP) components both production-induced imperfections such as fiber misalignment or pores and usage induced damages such as delamination caused by e.g. low velocity impacts or matrix and fiber failure may occur. Such imperfection and damages may lead to a reduction in the load bearing capabilities where early failure detection could be essential for the prognostic of future behaviour of the structure. Within this paper, the authors present the latest results on on-line Acoustic Emission monitoring of different pre-damaged composite coupons. Therefore, composite panels with different amounts of porosity and different impact damages were produced. The pre-damaged composite samples were subjected to a number of interrupted specific loads – bending, shear, tensile and tensile fatigue loads – to assess the Acoustic Emission signature in dependence of load and damage status. The damage status was further assessed by periodic computer tomography and ultrasonic inspection to compare the different NDT methods.

This study presents an approach for the detection of evolving degradation in large-scale low-speed roller bearings by clustering of Acoustic Emission (AE) events, and its application to experimental degradation data. To acquire the latter, a purpose-built linear bearing, representative of a segment of a turret bearing, has been instrumented with multiple piezoelectric AE transducers in the frequency range between 40–580 kHz. Clustering based on cross-correlation has identified a number of significant clusters that are linked to the observed damage. The results suggest that condition monitoring based on AE waveform similarity clustering is suitable for detection and identification of degradation in a large-scale roller bearing.

In this work, the influence of environmental conditions and damage on structures with closely spaced modes is investigated. Closely spaced bending modes are typical for symmetrical structures, such as towers. When monitoring such structures, identifying the mode shapes in particular is a challenge. In addition to environmental effects, the identification uncertainty is elevated for closely spaced modes, which leads to higher scatter in the identified mode shapes.In order to be able to investigate and quantify the influences, data from a structure with repairable damage is used. In particular, the investigation is carried out on long-term measurement data of the Leibniz University Test Structure for Monitoring (LUMO). LUMO enables the examination of different damage positions under ambient excitation and environmental conditions. For the identification, Bayesian Operational Modal Analysis (BAYOMA) is used.By considering the uncertainty of the modal parameters, especially for the mode shapes, a better understanding of the influence of damage and environmental conditions on the structure can be obtained. This is helpful for further monitoring stages, such as damage localisation using modal updating.

This work is focused on the design and manufacturing of an adaptive sensorized “Human Touch” membrane for medical bandages monitoring. The designed technological demonstrator consists of a polyurethane elastomeric shell with a thin composite supporting core (glass reinforced fibers and epoxy matrix). Fiber Bragg Grating Sensors (FBGS) embedded within this core allow to detect blood pressure with the aim of adjusting the tightness of the bandage so that it does not block the blood flow of the patient. At each FBG sensor, the membrane has a protrusion (a “button”) which amplifies the pressure on the sensor applying a longitudinal deformation to the fiber itself. To achieve this effect, the optical fiber must also be supported by two rigid constraints at both sides of the FBG sensor by inserting the composite core provided by circular holes in correspondence with the elastomeric button (and therefore the sensor). This core provides the additional function of obtaining the membrane stiffness necessary to guarantee a mutual decoupling in the presence of several FBG sensors within the membrane. After having described the main steps of the technological process developed for the manufacturing of the device, the subsequent calibration and validation tests carried out are reported in the second part of this memory. In particular, the membrane was applied to the bandage of an arm to evaluate the ability of the device to monitor both the bandaging action of the leg, giving feedback to the health worker and the period of maintenance thereof to avoid incorrect bandages.

Although relevant examples of systems devoted to shape sensing, damage detection, load identification, etc., do exist, even based on fiber optic sensors, they are barely suited for installation on real aerospace applications due to important criticalities in the application of the fiber.However, all HUMS application based on fiber optics can be enabled thanks to the proposed Smart Veil: a technology consisting of a thin composite membrane incorporating fiber optics placed on a complex path. This integrated element makes the monitoring system easy to use/handle, robust and reliable in real operating scenarios, capable of guaranteeing precise measurements.The effectiveness of this technological solution is proven by means of the manufacturing of a sensorized helicopter blade mockup. Among the several techniques that can be adopted using a fiber optic, FBG (Fiber Bragg Grating) sensors have been selected, allowing a robust punctual measure in specific locations.The choice of sensors position was led by the idea to exploit a digital twin for shape reconstruction and load identification, so all the components of the relevant loads acting on a tail rotor blade during operation can be obtained: axial, beam bending, chord bending and torsion. All sensors benefit from the use of the smart veil that guarantees robust and precise measures. Particularly, the torsional sensors do, since they need to be placed on an ad hoc path.The static calibration test and a comparison with a strain gauges system permitted a validation and showed the advantages of the technological solution proposed.

The aim is to create a smart structure holding the benefits of a structural monitoring system combined with the high performance of composite materials in a key component of a racing motorcycle, that is the front fork.Telescopic forks in composite material allow mass reduction while maintaining adequate stiffness, even if their use is currently limited only to a niche of applications, mainly for wear issues. Here, the fork tube was designed, optimizing the lamination sequence to replicate the stiffness of the original fork. The applied monitoring system uses optical fibers, for which composites are an excellent host material, with Bragg grating sensors. From the deformations measured is possible to derive information on the integrity, to set the performance of stiffness and to detect the loads.Given the difficulties in the correct positioning of the sensors (diametrically opposite on a cylindrical surface) and to imagine an effective use during racing, the need was highlighted to implement a system allowing a safe, easy, and precise positioning of the optical fiber on a predetermined path between the composite layers and the proper housing of the relative connector.The fork tube equipped with the smart veil was made by autoclave forming and then tested in the laboratory both alone and mounted on the bike. The result is a component whose deformations are known in real time, that integrates adequately protected and well-placed sensors, that are linked to the acquisition by means of a connector embedded directly into the structure itself.

Tensegrity is a network of bars and cables that maintains its structural integrity with tension present in its cables. Other than typical structural failure mechanisms, tensegrity may fail due to slacking of cables or buckling of bars. Real-life tensegrities are an assemblage of component modules. Large tensegrities require excessive computation for model-based structural health monitoring (SHM), which may sometimes make the problem ill-posed. Instead of the entire domain, only a substructure can be investigated explicitly. Substructures decouple the structure into independent components that can be monitored individually, provided the sub-domain interface is measured. Yet the integration of substructures within predictor-corrector model-based SHM algorithms needs special investigation from consistency, stability, and accuracy perspectives. To consider system uncertainties Bayesian filtering-based SHM approaches have been employed in this study. The need for interface measurement has been circumvented through an output injection approach. To increase computational efficiency, the domain decomposition approach is coupled with an interacting filtering-based approach that employs Ensemble Kalman filter (for state estimation) within an envelope of Particle filter (for health parameter estimation). This facilitates simultaneous estimation of state and parameters while enabling full parallelization capability.The proposed approach is tested on a six-stage tensegrity tower made of component simplex modules.

This paper proposes a Phononic crystal (PnC) device formed by embedding spheres in a polymer matrix for energy harvesting, an application that has drawn much research interest recently. A parametric study is performed to evaluate the optimal radius to lattice parameter ratio r/a = 0.4 to obtain a large bandgap as well as gap to mid-gap ratio. By removing a sphere from the central lattice point, a defect is created within an otherwise perfect PnC. The supercell technique is employed to compute defect modes and band structure of defect PnC. These defect bands show strong vibrational modes within the spatial arrangement of the defect point and these defect frequencies appear to be flat (dω/dk = 0) in the band structure confirming energy localization at these modes. Also, while varying the radius of the defect sphere rd in the range of [0.1a–0.5a], the number of defect modes within the bandgap tend to decrease, vanishing when rd equals 0.4a and 0.5a. To utilize the defect mode for energy harvesting, a small patch made of a piezoelectric material is attached onto the PnC device exactly above the defect. Introduction of this patch renders the band structure and defect modes to shift upwards compared to the defect PnC without the patch. A highly localized vibrational eigen mode is observed around 662 kHz in the band structure for the defect PnC with patch. The voltage response of this device is simulated by connecting a high resistance (open-circuit) to the patch and sweeping frequencies around the defect mode. This process produced a voltage of 2.8 V at 656.7 kHz for a 0.3 µm displacement excitation, thus demonstrating the usefulness of the proposed concept.

This paper presents recent research into low-cost flow meter designs to measure dripping leaks for health monitoring of water equipment. Water handling equipment is pervasive in our society, to the point it is often overlooked; including plumbing, faucets, bathtubs, showers, toilets, water heaters, water softeners, filtration systems, dish washers, washing machines, etc. Monitoring of the health and usage of water equipment will be a major contribution to the internet of things, providing information about equipment’s need for maintenance or re-placement to both conserve water and prevent costly flood damage.Detection of leaks and flow rates are core to monitoring the condition of water equipment. However, commercially available flow meters capable of measuring flow on the order of a dripping leak are too expensive for mass deployment in consumer products (>$300). This paper presents parametric studies of low cost in-line flow meter designs, intended to measure extreme low flow rates (<50 ml/min on the order of a dripping leak in ½ inch pipe). Designs were created based on a hybrid of traditional orifice and target meters. Part counts were kept minimal and synthesis methods were kept simple with the goal of minimizing cost.Parametric studies were performed to evaluate the effects of geometric dimensions on the sensitivity, pressure drop, and other traits of the flow meters. Testing was performed both with custom equipment and in the Badger Meter Flow Lab at the Global Water Center in Milwaukee Wisconsin. Results were compared to electromagnetic and mass-based measurements.

The analysis of dynamic systems via statistical models built on the measured data is the core idea of the so–called System Identification (SysId), i.e. a methodology which has proven to be one of the most effective tools for spectral analysis and, by extension, for the analysis of vibrating structures. In practical scenarios, where the operational and environmental uncertainties might be responsible for significant, additive noise levels, models more sophisticated than the standard ones are required. This requirement can be efficiently pursued via Errors–In–Variables (EIV) systems, which offer a powerful means to precisely tackling inherent disturbances hidden within real signals.In this work, we overcome the limitations of the classical implementation of output–only EIV models in the time domain by adopting a novel frequency domain approach defined in the Frisch scheme to solve, in a more direct way, the trade–off between frequency resolution and noise level. When applied for vibration diagnostics of a large-scale structure (i.e., a wind turbine blade prototype), the novel EIV–based SysId strategy proved significant accuracy in tracking frequency–related structural changes. Thus, it can be considered as a promising strategy for vibration assessment thanks to its reduced computational complexity and the quality of the retrieved spectral signature.

Structural health monitoring is a fundamental task to evaluate the condition of the structure and ensure an early detection of any material failure. This is important especially in highly stressed parts of a structure. In turbomachinery, it is in particular a rotating blade that must withstand the action of superheated steam flow. Excessive tension acting in a blade can then decrease its residual life that indicates the structural health of the blade.For the purpose of structural health monitoring of the rotating blades the high cycle fatigue (HCF) analysis tool was developed and is introduced in this paper. The HCF module is based on data processing of signals acquired using blade tip timing (BTT) system and provides the residual life of the blades in several critical points, defined along the blade body.The introduced HCF module was integrated into remote monitoring system (RMS) developed by the authors of this paper. The HCF module together with RMS is described in this paper in detail. The benefits of remote sensing are highlighted as well. Until today, tens of power plants worldwide are involved in presented RMS.

Over their life, space equipment needs to withstand strong high-frequency shocks, which could cause mission and safety critical damages. In order to verify the compliance with safety standards, pyroshock tests are employed. Based on launch vehicle characteristics, the requirements for the qualification of space equipment are usually established following the NASA-STD-7003A international standards in terms of a Shock Response Spectrum (SRS) representing the damage potential of the shock. Laboratory tests should then match the actual stress conditions reached during a real launch. Historically, this was obtained by means of explosive charges (hence the name “pyroshock”). Nevertheless, to foster repeatability and safety in laboratories, hammers or bullets are commonly used in nowadays shock testing machines.In this work, a resonant fixture test bench is considered. In this very common layout, a resonant metallic plate is interposed between the impact location and the test component so as to better simulate the shocks. The response of the resonant plate - which determines the required shock response spectrum - is currently empirically tuned by adding masses, damping, stiffness, or by varying the nature of the impact. This study aimed at developing a numerical model able to completely simulate a pyroshock test. Such a model can be used both for designing and for tuning the test bench so as to easily match different SRS requirements for different components under test. This leads to great economical advantages as can cut the calibration times leading to more efficient and effective testing.

Extreme–edge computing is becoming increasingly appealing for Structural Health Monitoring applications because it allows the optimal management of the available processing resources. This fosters the possibility to enhance the responsiveness and power management of the inspection system. To this end, particular attention should be warranted to the extreme–edge implementation of system identification (SysId) algorithms, which represent one of the most powerful means for dynamics analysis and, consequently, for vibration–based structural assessment.However, to implement this near–sensor processing, the design of an optimized hardware is fundamental. In this work, we fulfil this goal by proposing a novel smart accelerometer node, which is built on the combination of a wireless communication module, a high–performance microcontroller unit (MCU) and two tri-axial MEMS accelerometers, necessary to efficiently trigger the acquisition on energy thresholds while maximizing the energy saving. In normal operating mode, the MCU benefits from a clock frequency up to 80 MHz, Digital Signal Processing functionalities and a Floating Point Unit which make feasible the computation of SysId techniques at the extreme edge. Tested on a laboratory steel beam under varying damage levels, the developed system showed promising accuracy in tracking the variations of the vibration response of the structure.

Three-dimensional (3D) construction printing is an emerging alternative to conventional construction methods. Common gantry- and robotic arm-based systems impose scalability limitations based on the printer size. Several research efforts proposed using multiple mobile 3D printers towards large-scale printing, relying however on unrealistic assumptions of continuous communication among all agents. Here, we explore an active sensing framework allowing individual agents to assess other agents’ progress without directly communicating with them. Our approach leverages environmental modifications introduced by each agent during printing to track the structure evolution. We focus on heat conduction in the structure, which we discretize as a 2D lattice embodying its topology. Using on-board sensors, agents measure temperature and heat at their location, which they use to infer structure’s topology. From the input-output time-series and prior knowledge of the printing task, an agent identifies the system state-matrix by using a subspace identification method and solving an inverse eigenvalue problem. We demonstrate the validity of our approach through numerical simulations, establishing conditions for successful inference. We highlight the potential of the framework in facilitating information flow among agents through the physical medium, paving the way for decentralized collective mobile 3D printing.

Inferring the size of a complex system from partial measurements of some of its units is a common problem in engineering, with significant applications in the field of structural health monitoring (SHM), where one may attempt at relating system size (number of degrees of freedom) to the integrity of the structure. Here, we demonstrate the possibility of inferring the size of a stochastic system by assembling measurements of its response into a detection matrix. In deterministic systems, the rank of the detection matrix (number of non-zero singular values) equals the size of the largest observable system component. We extend this framework to reconstruct the number of states of an unknown Markov chain, where we cannot distinguish between two or more states. In this case, we only have access to an estimate of the detection matrix, but with a larger rank, since stochasticity generates a series of non-zero singular values. We establish conditions for the correct inference of system size, relating the number of realizations and the smallest true singular value. Our work highlights connections between SHM, system identification, and control theory, paving the way for new cross-disciplinary inquiries.

Aging infrastructure is increasingly becoming a problem in industrialized countries. As a matter of fact, many existing bridges were already built in the mid-to-late twentieth century and hence are not in a good condition. A majority of these bridges were built from prestressed concrete and therefore may suffer from stress corrosion cracking of steel tendons. Due to their importance for structural integrity and hence structural safety, a reliable monitoring method of bridge tendons is of major interest. Recent research suggests the use of acoustic emissions (AE) using ultrasonic sensors to detect wire break events in civil engineering structures. In this work, accelerometers operating in the audible frequency range (<17 kHz) are used to record more than hundred wire breaks manually conducted at two old girders taken from a demolished bridge in Roding, Germany. Based on these experiments, approximately twelve days of continuous operational recordings conducted at a similar bridge construction in Hagen, Germany and additional wire breaks recorded at the bridge structure in Roding before demolition, we train and evaluate a support vector machine to detect wire break events in bridge tendons using low-level frequency domain features such as the spectral bandwidth. On the evaluation dataset, we achieve a recall score of more than 90% while having a FP-rate of approximately ten events per day. It should be noted that using low-frequency acoustic emissions we are able to detect wire break events in distances up to more than 20 m.

App4SHM is a smartphone application for structural health monitoring (SHM). It can be applied to perform SHM of bridges or other special structures to assess their condition after a catastrophic event or when required by authorities. The application interrogates the phone’s internal accelerometer to measure accelerations, then estimates the natural frequencies, and compares them with a reference data set through a machine learning algorithm properly trained to detect damage. The machine learning is fundamental to take into account the effects of operational and environmental variability on the damage detection. A server is accessed and used by the application to run most of the computational operations and store the data sets. As a customized SHM process, App4SHM follows four main steps: (i) structure identification; (ii) data acquisition; (iii) feature extraction, which calls the server to estimate the first three natural frequencies and stores them into a feature vector (observation); and (iv) damage detection, where a damage indicator is computed for each new observation, based on the Mahalanobis-squared distance. The damage indicator of the new observation is plotted, and a flag is raised green if the structure is undamaged and raised red if structural damage is suspected. To test the robustness of the application, the damage detection capability was tested on real data sets from two twin post-tensioned concrete bridges in Brazil under traffic and temperature variability. The natural frequencies obtained from the application were also compared with the ones estimated using data sets from a traditional data acquisition system.

Offshore platforms can be subjected to harsh environmental conditions increasing their vulnerability to failures. Hence, the deployment of a Structural Health Monitoring (SHM) system is highly relevant to safeguard their structural integrity. Typically, an SHM system includes a Damage Detection (DD) scheme, which aims to identify structural damages by using information recorded by the sensing system. Plenty of DD schemes have been proposed for various types of structures. However, several peculiarities of the offshore platforms, being associated with the near impossibility to operate sensors underwater, can challenge the performance of conventional DD schemes. In this context, the current study emphasizes the development of a new Kalman filter-based framework that is anticipated to estimate the location and severity of the damage. Especially, the DD problem is approached as an input-state estimation problem for nonlinear systems. The main objective of this method is to identify and quantify the damages induced by wave loads in terms of, for example, plastic deformations. Advanced modeling techniques are required for this task. The proposed DD scheme is tested on a 2D steel jacket structure subjected to wave loading. The outcome of the specific DD scheme is expected to provide the damage profile of the structure after various hazardous events.

Structural performance prediction of innovative connection details requires both advanced design tools and analysis models that are verified through experimental data. For most civil structures, field or full-scale tests to failure are not feasible, therefore scale model laboratory experiments are critical to advancing the state-of-the-art for structural design. To isolate the target structural behavior for evaluation, all structural behaviors, including impact of boundary conditions and members interactions, must be controlled and accurately accounted for in the associated structural models.Data collection including installation procedure, sensors type, sampling rate and data acquisition system are critical components to any experimental or field fatigue test. Fatigue experiments often benefit from data collected at locations that are difficult, if not impossible, to instrument with traditional contact strain gauges, such as weld toes or curved members. This paper presents the experimental fatigue evaluation of a unique steel connection, including the instrumentation plan, which utilizes contact and non-contact measurements, and collected data from a high cycle fatigue experiment of a scale model specimen of a gusset-less truss bridge connection. The monitoring protocol for an experimental fatigue test setup and the evaluation of structural response measurements is developed, implemented, and tested as part of this study. This paper also presents the use of numerical modeling to supplement experimental testing, and the high-cycle fatigue testing results of the novel connection studied herein.