Nothing Special   »   [go: up one dir, main page]
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
7.3
3.4
Journals
Sensors
Special Issues
Advanced Sensing Technologies in Structural Health Monitoring and Its Applications
sensors-logo
Submit to Special Issue Submit Abstract to Special Issue Review for Sensors Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Advanced Sensing Technologies in Structural Health Monitoring and Its Applications

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Sensor Materials".

Deadline for manuscript submissions: 20 April 2025 | Viewed by 55443

Special Issue Editor

Prof. Dr. Ricardo Perera

E-Mail Website
Guest Editor
Department of Mechanical Engineering, Technical University of Madrid, 28006 Madrid, Spain
Interests: structural health monitoring; electromechanical impedance method; PZT; FBG

Special Issue Information

Dear Colleagues,

Today, structural health monitoring (SHM) is an important research area because of its strong connection with structural safety and the need to monitor and extend the lives of existing structures. Recent years have shown a rapid development of different technologies and sensing techniques developed for structural monitoring. Based on the data extracted from these technologies, SHM algorithms are used to give information and make decisions about structural conditions.

The rapid development of advanced sensing technologies will overcome the challenging issues in the realization of smart systems and structures.

The aim of this Special Issue is to focus on the most recent strategies and development of innovative sensors and biosensors, as well as their applications for structural monitoring.

Both review articles and original research papers are welcome.

Prof. Dr. Ricardo Perera
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • structural health monitoring (SHM)
  • condition monitoring
  • sensing technologies
  • advanced sensors
  • smart systems and structures
  • data processing
  • artificial intelligence
  • machine learning

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue polices can be found here.

Published Papers (28 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

Jump to: Review

20 pages, 3887 KiB  
Article
Effect of Damping on the Identification of Bridge Properties Using Vehicle Scanning Methods
by Emrah Erduran and Semih Gonen
Sensors 2024, 24(17), 5785; https://doi.org/10.3390/s24175785 - 5 Sep 2024
Viewed by 693
Abstract
Vehicle scanning methods are gaining popularity because of their ability to identify modal properties of several bridges with only one instrumentation setup, and several methods have been proposed in the last decade. In the numerical models used to develop and validate such methods, [...] Read more.
Vehicle scanning methods are gaining popularity because of their ability to identify modal properties of several bridges with only one instrumentation setup, and several methods have been proposed in the last decade. In the numerical models used to develop and validate such methods, bridge damping is often overlooked, and its impact on the efficacy of vehicle scanning methods remains unknown. The present article addresses this knowledge gap by systematically investigating the effects of bridge damping on the efficacy of vehicle scanning methods in identifying the modal properties of bridges. For this, acceleration responses obtained from a numerical model of a bridge and vehicle are used. Four different scenarios are considered where vehicle damping, presence of road roughness, and traffic on the bridge are varied. Bridge damping is modeled using mass-proportional, stiffness-proportional, and Rayleigh damping models. The impacts of ignoring bridge damping or considering one of these damping models on the modal frequencies and mode shapes identified using the vehicle response are investigated by comparing the results. The outcomes of the numerical analysis show that ignoring bridge damping in vehicle scanning applications can significantly increase the efficacy of these methods. They also show that the identifiability of the bridge frequencies and bridge mode shapes from the vehicle response decreases significantly when bridge damping is considered. Further, the damping model used impacts which bridge modes can be identified because different damping models provide different modal damping ratios for each mode. The results highlight the importance of correctly simulating damping behavior of bridges, which is often ignored, to be able to correctly evaluate the efficacy of vehicle scanning methods, and they provide an important stepping stone for future studies in this field. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Mathematical model consisting of the bridge and the vehicle.</p> Full article ">Figure 2
<p>Road roughness profile.</p> Full article ">Figure 3
<p>Variation of modal damping ratio with frequency.</p> Full article ">Figure 4
<p>Fourier amplitude spectrum of the vibrations at the quarter-span of the bridge for different damping models.</p> Full article ">Figure 5
<p>Acceleration response of the vehicle for different damping models.</p> Full article ">Figure 6
<p>Contact point accelerations for different damping models for the undamped vehicle.</p> Full article ">Figure 7
<p>Fourier amplitude spectrum of the contact points for different damping models for the undamped vehicle.</p> Full article ">Figure 8
<p>Fourier amplitude spectrum of the contact point response of the undamped vehicle for different bridge damping models considering road roughness class A.</p> Full article ">Figure 9
<p>Relative energy at the first three modal frequencies in the contact point response for different vehicle damping values and bridge damping models considering road roughness class A.</p> Full article ">Figure 10
<p>FAS of the contact point response for different bridge damping models considering road roughness class A and traffic.</p> Full article ">Figure 11
<p>First three modal components identified from the contact point response for smooth road profile.</p> Full article ">Figure 12
<p>Comparison of first three mode shapes identified from the contact point response with the analytical solution for smooth road profile.</p> Full article ">Figure 13
<p>Mode shapes identified from the contact point response of the vehicle travelling on rough road.</p> Full article ">Figure 14
<p>Mode shapes identified from the contact point response of the vehicle travelling on rough road considering vibrations from existing traffic.</p> Full article ">
12 pages, 3454 KiB  
Article
Accurate Ultrasonic Thickness Measurement for Arbitrary Time-Variant Thermal Profile
by Rajendra P. Palanisamy, Do-Kyung Pyun and Alp T. Findikoglu
Sensors 2024, 24(16), 5304; https://doi.org/10.3390/s24165304 - 16 Aug 2024
Viewed by 828
Abstract
Ultrasonic thickness measurement of mechanical structures is one of the most popular and commonly used nondestructive methods for various kinds of process control and corrosion monitoring. With ultrasonic propagation speed being temperature-dependent, the thickness measurement can be performed reliably only when the thermal [...] Read more.
Ultrasonic thickness measurement of mechanical structures is one of the most popular and commonly used nondestructive methods for various kinds of process control and corrosion monitoring. With ultrasonic propagation speed being temperature-dependent, the thickness measurement can be performed reliably only when the thermal profile is completely known. Most conventional techniques assume the temperature of the test structure is uniform and at room temperature across its thickness. Such assumptions may lead to large errors in the thickness measurement, especially when there are significant temperature variations across the thickness. State-of-the-art techniques use external temperature measurements or implement iterative methods to compensate for the unknown thermal profiles. However, such techniques produce unsatisfactory results when the heat distribution is complex or varies rapidly with time. In this work, we propose a two-sensors technique, using both compressive and shear excitations, with a non-iterative rapid data processing method for accurate thickness measurement under arbitrary time-variant thermal profile. The independent behavior of shear and compressive waves is used to formulate a real-time thickness estimation technique. The developed technique is experimentally validated on a steel plate with fixed acoustic sensors. Test results show that the error in thickness estimation can be reduced by up to 98% compared to conventional thickness gauging methods. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) Top view of the test plate with attached acoustic shear (S) and compressive (C) and thermocouple sensors. (<b>b</b>) Test plate along with attached sensors placed in the oven for calibration.</p> Full article ">Figure 2
<p>Schematic of the experimental setup to perform sequential pulse-echo measurement using shear (S) and compressive (C) acoustic sensors.</p> Full article ">Figure 3
<p>(<b>a</b>) Temperature measured at the top and bottom of the test plate during calibration. (<b>b</b>) Calculated compressive and shear speed in the test plate during calibration.</p> Full article ">Figure 4
<p>Acquired signal during pulse-echo measurement with (<b>a</b>) compressive and (<b>b</b>) shear acoustic sensors. The first and fourth echo used for TOF calculation are highlighted.</p> Full article ">Figure 5
<p>(<b>a</b>) Measured linear relationship between the temperature and the speed of shear and compressive modes. (<b>b</b>) Calculated linear relationship between the temperature and the speed ratio.</p> Full article ">Figure 6
<p>Schematic of test plate with arbitrary thermal profile T(x).</p> Full article ">Figure 7
<p>Schematic of experimental test setup (front and side cross-section view) that illustrates the placement of heating tape, acoustic sensor, and thermocouple. Red shading is used to indicate the existence of temperature gradients during validation experiments.</p> Full article ">Figure 8
<p>(<b>a</b>) Measured TOF for compressive and shear sensors under continuous heating. (<b>b</b>) Comparing measured temperature using thermocouple attached to the outer plate and indirect average temperature predicted via the two-sensor method under continuous heating.</p> Full article ">Figure 9
<p>(<b>a</b>) Thickness measurement error for all three methods shown under continuous heating. (<b>b</b>) Measurement errors for the ‘1-sensor + temp’ method and the proposed ‘2-sensor’ methods, separately compared for better visualization.</p> Full article ">Figure 10
<p>(<b>a</b>) Measured TOF for compressive and shear sensors under intermittent heating. (<b>b</b>) Comparing measured temperature using thermocouple attached to the outer plate and the indirect average temperature predicted by the two-sensor method under intermittent heating.</p> Full article ">Figure 11
<p>(<b>a</b>) Thickness measurement errors for all three methods shown under intermittent heating. (<b>b</b>) Measurement errors for the ‘1 sensor + temp’ method and the proposed ‘2-sensor’ method are separately compared for better visualization.</p> Full article ">
25 pages, 7064 KiB  
Article
Structural Damage Detection Based on the Correlation of Variational Autoencoder Neural Networks Using Limited Sensors
by Jun Lin and Hongwei Ma
Sensors 2024, 24(8), 2616; https://doi.org/10.3390/s24082616 - 19 Apr 2024
Viewed by 995
Abstract
Identifying the structural state without baseline data is an important engineering problem in the field of structural health monitoring, which is crucial for assessing the safety condition of structures. In the context of limited accelerometers available, this paper proposes a correlation-based damage identification [...] Read more.
Identifying the structural state without baseline data is an important engineering problem in the field of structural health monitoring, which is crucial for assessing the safety condition of structures. In the context of limited accelerometers available, this paper proposes a correlation-based damage identification method using Variational Autoencoder neural networks. The approach involves initially constructing a Variational Autoencoder network model for bridge damage detection, optimizing parameters such as loss functions and learning rates for the model, and ultimately utilizing response data from limited sensors for model training analysis to determine the structural state. The contribution of this paper lies in the ability to identify structural damage without baseline data using response data from a small number of sensors, reducing sensor costs and enhancing practical applications in engineering. The effectiveness of the proposed method is demonstrated through numerical simulations and experimental structures. The results show that the method can identify the location of damage under different damage conditions, exhibits strong robustness in detecting multiple damages, and further enhances the accuracy of identifying bridge structures. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Autoencoder Neural Network Architecture Diagram.</p> Full article ">Figure 2
<p>Internal operation of VAE.</p> Full article ">Figure 3
<p>The structure of VAE.</p> Full article ">Figure 4
<p>Data augmentation process.</p> Full article ">Figure 5
<p>T Simply Supported Beam Model.</p> Full article ">Figure 6
<p>Fourier spectrum of the acceleration response.</p> Full article ">Figure 7
<p>Damage factor curves in healthy status.</p> Full article ">Figure 8
<p>Damage factor curves under different damage severities with mass moving at 0.2 m/s: (<b>a</b>) γ = 10%, (<b>b</b>) γ = 10% to γ = 50% (The red triangle marks the location of the damage).</p> Full article ">Figure 9
<p>Damage factor curves under different damage severities with mass moving at 0.5 m/s: (<b>a</b>) γ = 20%, (<b>b</b>) γ = 10% to γ = 50% (The red triangle marks the location of the damage).</p> Full article ">Figure 10
<p>Damage factor curves under different damage severities with mass moving at 0.7 L damage location and 0.2 m/s: (<b>a</b>) γ = 10%, (<b>b</b>) γ = 10% to γ = 50% (The red triangle marks the location of the damage).</p> Full article ">Figure 11
<p>Damage factor curves for single damage detection under different noise: (<b>a</b>) γ = 10% to 50%, SNR =40 dB, (<b>b</b>) γ = 10% to 50%, SNR = 25 dB (The red triangle marks the location of the damage).</p> Full article ">Figure 12
<p>Damage factor curves for multiple damage detection under different damage severities: (<b>a</b>) MDC1, (<b>b</b>) MDC3, (<b>c</b>) MDC1 to MDC5 (The red triangle marks the location of the damage).</p> Full article ">Figure 13
<p>Damage factor curves for multiple damage detection under three damage conditions (The red triangle marks the location of the damage).</p> Full article ">Figure 14
<p>Experimental setup: (<b>a</b>) Experimental structure, (<b>b</b>) constant-speed bridges.</p> Full article ">Figure 15
<p>Experimental setup: (<b>a</b>) model car, (<b>b</b>) DH5920 dynamic test system.</p> Full article ">Figure 16
<p>Experimental diagram.</p> Full article ">Figure 17
<p>Vertical view of experimental damage: (<b>a</b>) side view of the damage fracture, (<b>b</b>) top view of the bottom damage.</p> Full article ">Figure 18
<p>Acceleration data of experiment (The red arrow marks the location of the damage).</p> Full article ">Figure 19
<p>Damage factor curves calculated from the acceleration responses in healthy status.</p> Full article ">Figure 20
<p>Damage factor curves for single damage detection under different damage severities: (<b>a</b>) mass with 10.5 kg and the speed with 0.25 m/s, (<b>b</b>) mass with 20.5 kg and speed with 0.25 m/s (The red triangle marks the location of the damage).</p> Full article ">Figure 21
<p>Damage factor curves for single damage detection under different damage severities: (<b>a</b>) mass with 10.5 kg and the speed with 0.5 m/s, (<b>b</b>) mass with 20.5 kg and speed with 0.5 m/s (The red triangle marks the location of the damage).</p> Full article ">Figure 22
<p>Damage factor curves for mutiple damage detection under different damage severities: (<b>a</b>) mass with 10.5 kg and the speed with 0.25 m/s, (<b>b</b>) mass with 20.5 kg and speed with 0.25 m/s (The red triangle marks the location of the damage).</p> Full article ">Figure 23
<p>Damage factor curves for mutiple damage detection under different damage severities: (<b>a</b>) mass with 10.5 kg and the speed with 0.5 m/s, (<b>b</b>) mass with 20.5 kg and speed with 0.5 m/s (The red triangle marks the location of the damage).</p> Full article ">
18 pages, 7108 KiB  
Article
Effect of Sonication Batch on Electrical Properties of Graphitic-Based PVDF-HFP Strain Sensors for Use in Health Monitoring
by Victor Díaz-Mena, Xoan F. Sánchez-Romate, María Sánchez and Alejandro Ureña
Sensors 2024, 24(6), 2007; https://doi.org/10.3390/s24062007 - 21 Mar 2024
Cited by 2 | Viewed by 1438
Abstract
In this study, flexible nanocomposites made from PVDF-HFP reinforced with carbon nanotubes (CNTs) and graphene nanoplatelets (GNPs) are manufactured using a sonication and solvent casting method for monitoring purposes. More specifically, the effect of the volume batch under the sonication process is explored. [...] Read more.
In this study, flexible nanocomposites made from PVDF-HFP reinforced with carbon nanotubes (CNTs) and graphene nanoplatelets (GNPs) are manufactured using a sonication and solvent casting method for monitoring purposes. More specifically, the effect of the volume batch under the sonication process is explored. For CNT-based composites, the electrical conductivity decreases as the batch volume increases due to less effective dispersion of the CNTs during the 30-min sonication. The maximum electrical conductivity achieved in this type of sensor is 1.44 ± 0.17 S/m. For the GNP-based nanocomposites, the lower the batch volume is, the more breakage of nanoplatelets is induced by sonication, and the electrical response decreases. This is also validated by AC analysis, where the characteristic frequencies are extracted. Here, the maximum electrical conductivity measured is 8.66 ± 1.76 S/m. The electromechanical results also show dependency on the batch volume. In the CNT-based nanocomposites, the higher gauge factor achieved corresponds to the batch size, where the sonication may be more effective because it leads to a dispersed pathway formed by aggregates connected by tunneling mechanisms. In contrast, in the CNT-based nanocomposites, the GF depends on the lateral size of the GNPs. The biggest GF of all sensors is achieved with the PVDF-HFP/GNP sensors, having a value of 69.36 × 104 at 35% of strain, while the highest GF achieved with a PVDF-HFP/CNT sensor is 79.70 × 103 at 70%. In addition, cycling tests show robust electromechanical response with cycling for two different strain percentages for each type of nanocomposite. The sensor with the highest sensitivity is selected for monitoring two joint movements as proof of the applicability of the sensors manufactured. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Manufacture and characterization routes for the two types of sensors manufactured. Steps 1 and 2 show the steps followed in manufacturing the thin and highly stretchable nanocomposite films and steps 3 and 4 show the microstructural and electromechanical characterization of the sensors.</p> Full article ">Figure 2
<p>TEM images of the CNTs after 30 min of sonication in (<b>a</b>) 10 mL, (<b>b</b>) 20 mL, (<b>c</b>) 30 mL, and (<b>d</b>) 40 mL batch volumes of PVDF-HFP dissolved in DMF.</p> Full article ">Figure 3
<p>SEM images of the GNPs after 30 min of sonication in (<b>a</b>) 10 mL, (<b>b</b>) 20 mL, (<b>c</b>) 30 mL, and (<b>d</b>) 40 mL batch volumes of PVDF-HFP dissolved in DMF.</p> Full article ">Figure 4
<p>Electrical conductivities of CNT-based sensors (<b>left</b>) and GNP-based sensors (<b>right</b>) as a function of ultrasonication batch volumes.</p> Full article ">Figure 5
<p>Bode plots for all conditions manufactured: (<b>a</b>–<b>c</b>) the CNT-based composites and (<b>d</b>–<b>g</b>) the GNP-based composites.</p> Full article ">Figure 6
<p>Changes in the characteristic frequency (fc) as a function of the volume of the batch.</p> Full article ">Figure 7
<p>Explanation of the distinct parts of the electrical circuit used for modeling the AC behavior of the samples.</p> Full article ">Figure 8
<p>Bode plots for all conditions manufactured: (<b>a</b>–<b>c</b>) the CNT-based composites and (<b>d</b>–<b>g</b>) the GNP-based composites with electrical circuit modeling represented as a line in all graphs.</p> Full article ">Figure 9
<p>R<sub>tun</sub> and R<sub>int</sub> (<b>top</b>) and R<sub>int</sub>/R<sub>tun</sub> (<b>bottom</b>) analysis for CNT-based sensors (<b>a</b>) and GNP-based sensors (<b>b</b>).</p> Full article ">Figure 10
<p>Tensile stress curves with electrical response (ΔR/R<sub>0</sub>) for both CNT-based sensors (<b>left</b>) and GNP-based sensors (<b>right</b>).</p> Full article ">Figure 11
<p>Gauge factors (GF) for CNT-based sensors (<b>left</b>) and GNP-based sensors (<b>right</b>) with the highest GF values marked for each, and a detail of the GF achieves at low percentages of strain.</p> Full article ">Figure 12
<p>Cyclic tests for CNT-based sensors for both (<b>a</b>) 5% and (<b>b</b>) 10% of strain reached.</p> Full article ">Figure 13
<p>Cyclic tests for GNP-based sensors for both (<b>a</b>) 1% and (<b>b</b>) 2.5% of strain reached.</p> Full article ">Figure 14
<p>Wrist (<b>a</b>) and finger (<b>b</b>) motion monitoring at different degrees to prove the sensor’s applicability in health monitoring.</p> Full article ">
14 pages, 26928 KiB  
Article
Autonomous Sensor System for Low-Capacity Wind Turbine Blade Vibration Measurement
by Diego Muxica, Sebastian Rivera, Marcos E. Orchard, Constanza Ahumada, Francisco Jaramillo, Felipe Bravo, José M. Gutiérrez and Rodrigo Astroza
Sensors 2024, 24(6), 1733; https://doi.org/10.3390/s24061733 - 7 Mar 2024
Cited by 2 | Viewed by 1520
Abstract
This paper presents the design, implementation, and validation of an on-blade sensor system for remote vibration measurement for low-capacity wind turbines. The autonomous sensor system was deployed on three wind turbines, with one of them operating in harsh weather conditions in the far [...] Read more.
This paper presents the design, implementation, and validation of an on-blade sensor system for remote vibration measurement for low-capacity wind turbines. The autonomous sensor system was deployed on three wind turbines, with one of them operating in harsh weather conditions in the far south of Chile. The system recorded the acceleration response of the blades in the flapwise and edgewise directions, data that could be used for extracting the dynamic characteristics of the blades, information useful for damage diagnosis and prognosis. The proposed sensor system demonstrated reliable data acquisition and transmission from wind turbines in remote locations, proving the ability to create a fully autonomous system capable of recording data for monitoring and evaluating the state of health of wind turbine blades for extended periods without human intervention. The data collected by the sensor system presented in this study can serve as a foundation for developing vibration-based strategies for real-time structural health monitoring. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Hardware disposition overview of the embedded damage detection autonomous system for WT blades.</p> Full article ">Figure 2
<p>Overview of the nose hub and its components. They are distributed in such way that the hub maintains its balance while rotating. Most components are secured with bolts and a silicone-based glue to absorb vibrations.</p> Full article ">Figure 3
<p>The graph shows the charge flow for both configurations. The higher consumption was due to the Wi-Fi module being used for data transmission. This led to an increase of 67.1% compared with normal operation using a base station. The lower consumption band shows the system in sleep mode, reducing consumption by 84.9%.</p> Full article ">Figure 4
<p>Oscilloscope capture illustrating sequential activation of nine digital channels with 5 ms division intervals.</p> Full article ">Figure 5
<p>Histogram illustrating the timing of the conducted measurements.</p> Full article ">Figure 6
<p>Battery voltage over time. The graph displays measurements taken up until the 15th of May. In addition to the raw measurements, the graph also features a moving average and the expected values for the remaining period of the year.</p> Full article ">Figure 7
<p>Cloud cover categories in Aysen throughout the year from [<a href="#B31-sensors-24-01733" class="html-bibr">31</a>].</p> Full article ">Figure 8
<p>Low-capacity wind turbine installed in south Chile (WT1). (<b>a</b>) Image of the installed wind turbine. (<b>b</b>) Identified natural frequencies of three modes from mid-January to mid-March 2023.</p> Full article ">Figure 9
<p>Vibration measurement sets for blades running at 15 r/min. (<b>a</b>) Image of the wind turbine simulation apparatus. (<b>b</b>) Healthy blade’s vibration signals from the <span class="html-italic">z</span> axis (<b>top</b>) and <span class="html-italic">x</span> axis (<b>bottom</b>). (<b>c</b>) FFTs for the <span class="html-italic">z</span> axis signal (<b>left</b>) and for the <span class="html-italic">x</span> axis signal (<b>right</b>) for the blade with no damage. (<b>d</b>) Damaged blade’s vibration signals from the <span class="html-italic">z</span> axis (<b>top</b>) and <span class="html-italic">x</span> axis (<b>bottom</b>). (<b>e</b>) FFTs for the <span class="html-italic">z</span> axis signal (<b>left</b>) and for the <span class="html-italic">x</span> axis signal (<b>right</b>) for the damaged blade.</p> Full article ">Figure 10
<p>Acceleration measurements recorded on blades of WT2. (<b>a</b>) Image of the instrumented WT2. (<b>b</b>) The <span class="html-italic">x</span> axis acceleration of sensor A1 and <span class="html-italic">z</span> axis acceleration of sensors A2 and A3. (<b>c</b>) Zoomed-in version of signals in (<b>b</b>).</p> Full article ">
24 pages, 8868 KiB  
Article
Unmanned Aerial Vehicle-Based Structural Health Monitoring and Computer Vision-Aided Procedure for Seismic Safety Measures of Linear Infrastructures
by Luna Ngeljaratan, Elif Ecem Bas and Mohamed A. Moustafa
Sensors 2024, 24(5), 1450; https://doi.org/10.3390/s24051450 - 23 Feb 2024
Cited by 1 | Viewed by 1826
Abstract
Computer vision in the structural health monitoring (SHM) field has become popular, especially for processing unmanned aerial vehicle (UAV) data, but still has limitations both in experimental testing and in practical applications. Prior works have focused on UAV challenges and opportunities for the [...] Read more.
Computer vision in the structural health monitoring (SHM) field has become popular, especially for processing unmanned aerial vehicle (UAV) data, but still has limitations both in experimental testing and in practical applications. Prior works have focused on UAV challenges and opportunities for the vibration-based SHM of buildings or bridges, but practical and methodological gaps exist specifically for linear infrastructure systems such as pipelines. Since they are critical for the transportation of products and the transmission of energy, a feasibility study of UAV-based SHM for linear infrastructures is essential to ensuring their service continuity through an advanced SHM system. Thus, this study proposes a single UAV for the seismic monitoring and safety assessment of linear infrastructures along with their computer vision-aided procedures. The proposed procedures were implemented in a full-scale shake-table test of a natural gas pipeline assembly. The objectives were to explore the UAV potential for the seismic vibration monitoring of linear infrastructures with the aid of several computer vision algorithms and to investigate the impact of parameter selection for each algorithm on the matching accuracy. The procedure starts by adopting the Maximally Stable Extremal Region (MSER) method to extract covariant regions that remain similar through a certain threshold of image series. The feature of interest is then detected, extracted, and matched using the Speeded-Up Robust Features (SURF) and K-nearest Neighbor (KNN) algorithms. The Maximum Sample Consensus (MSAC) algorithm is applied for model fitting by maximizing the likelihood of the solution. The output of each algorithm is examined for correctness in matching pairs and accuracy, which is a highlight of this procedure, as no studies have ever investigated these properties. The raw data are corrected and scaled to generate displacement data. Finally, a structural safety assessment was performed using several system identification models. These procedures were first validated using an aluminum bar placed on an actuator and tested in three harmonic tests, and then an implementation case study on the pipeline shake-table tests was analyzed. The validation tests show good agreement between the UAV data and reference data. The shake-table test results also generate reasonable seismic performance and assess the pipeline seismic safety, demonstrating the feasibility of the proposed procedure and the prospect of UAV-based SHM for linear infrastructure monitoring. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Proposed computer vision procedures for UAV-based seismic SHM for linear infrastructures.</p> Full article ">Figure 2
<p>Selected features (<math display="inline"><semantics> <mrow> <msub> <mi>P</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>P</mi> <mn>2</mn> </msub> <mo>,</mo> <msub> <mi>P</mi> <mn>3</mn> </msub> <mo>,</mo> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>B</mi> <mi>G</mi> </mrow> </semantics></math>) from Test 1 (<b>a</b>), Test 2 (<b>b</b>), and Test 3 (<b>c</b>) with the setup on the simulator (<b>e</b>). An example of targeted feature matching with no errors (<b>d</b>) in Test 1.</p> Full article ">Figure 3
<p>Gray-level distribution and intensity.</p> Full article ">Figure 4
<p>Detected regions (<b>top</b>) and differences from reference image (<math display="inline"><semantics> <mrow> <msub> <mo>∆</mo> <mrow> <mi>r</mi> <mi>e</mi> <mi>g</mi> <mi>i</mi> <mi>o</mi> <mi>n</mi> <mi>s</mi> </mrow> </msub> </mrow> </semantics></math> %, <b>bot</b>.) with respect to MSER threshold delta variations (MSER TH).</p> Full article ">Figure 5
<p>Detected MSERs and correct pairs from SURF, KNN, and refined MSAC matching concerning threshold delta variations using reference and second images from Test 1.</p> Full article ">Figure 6
<p>Percentage of correct matches (accuracy (%)) based on SURF 64-D and 128-D and KNN threshold variations (KNN TH).</p> Full article ">Figure 7
<p>Correct pairs and matching accuracies with SURF and MSAC algorithms based on SURF 64-D and 128-D with MSER threshold delta variations (MSER TH).</p> Full article ">Figure 8
<p>Number of correct pair matches with their respective accuracies (%) based on MSAC threshold (MSAC TH) with MSER threshold delta variations (MSER TH).</p> Full article ">Figure 9
<p>Point matching pairs with their respective MSAC thresholds. The example is taken from Test 1 and shows selected points <math display="inline"><semantics> <mrow> <msub> <mi>P</mi> <mn>1</mn> </msub> <mo>,</mo> <mo> </mo> <msub> <mi>P</mi> <mn>2</mn> </msub> <mo>,</mo> <mo> </mo> <msub> <mi>P</mi> <mn>3</mn> </msub> </mrow> </semantics></math>, and <math display="inline"><semantics> <mrow> <mi>B</mi> <mi>G</mi> </mrow> </semantics></math> in the specimen area and unidentified points.</p> Full article ">Figure 10
<p>Point matching pairs from Tests 2 and 3 and selected points to measure displacement.</p> Full article ">Figure 11
<p>Displacement response results, <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="sans-serif-bold-italic">δ</mi> <mi mathvariant="bold-italic">x</mi> </msub> </mrow> </semantics></math>, from validation Tests 1, 2, and 3.</p> Full article ">Figure 12
<p>AR spectrum and natural frequency of specimen measured by validation Tests 1–3.</p> Full article ">Figure 13
<p>Seismic testing setup showing the UAV position during tests, pipeline position on the biaxial shake table, and selected points to generate the seismic response.</p> Full article ">Figure 14
<p>Computer vision algorithm results from pipeline test. (<b>a</b>) Feature of interests, (<b>b</b>) Detected MSER, (<b>c</b>) SURF and KNN matching, (<b>d</b>) Refined matching results using MSAC.</p> Full article ">Figure 15
<p>Pipeline seismic responses in lateral and biaxial directions.</p> Full article ">Figure 16
<p>Frequency response and stabilization plots of pipeline system in lateral and longitudinal directions.</p> Full article ">
21 pages, 4617 KiB  
Article
Ceramic Stress Sensor Based on Thick Film Piezo-Resistive Ink for Structural Applications
by Gabriele Bertagnoli, Mohammad Abbasi Gavarti and Mario Ferrara
Sensors 2024, 24(2), 599; https://doi.org/10.3390/s24020599 - 17 Jan 2024
Cited by 2 | Viewed by 1195
Abstract
This paper presents a ceramic stress sensor with the dimension of a coin, able to measure the compressive force (stress) applied to its two round faces. The sensor is designed and engineered to be embedded inside concrete or masonry structures, like bridges or [...] Read more.
This paper presents a ceramic stress sensor with the dimension of a coin, able to measure the compressive force (stress) applied to its two round faces. The sensor is designed and engineered to be embedded inside concrete or masonry structures, like bridges or buildings. It provides good accuracy, robustness, and simplicity of use at potentially low cost for large-scale applications in civil structures. Moreover, it can be calibrated temperature compensated, and it is inherently hermetic, ensuring the protection of sensitive elements from the external environment. It is, therefore, suitable for operating in harsh and dirty environments like civil constructions. The sensor directly measures the internal stress of the structure, exploiting the piezo resistivity of thick film ink based on ruthenium oxide. It is insensitive with respect to the stiffness of the embedding material and the variation of the surrounding material properties like concrete hardening, shrinkage, and creep as it decouples the two components of stress. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Stress sensor [<a href="#B21-sensors-24-00599" class="html-bibr">21</a>].</p> Full article ">Figure 2
<p>Layers inside sensor: (<b>a</b>) Bottom ceramic layer (thickness 1.5 ÷ 2.0 mm); (<b>b</b>) Bottom glass frit—between bottom and intermediate ceramic layer (thickness 0.010 ÷ 0.020 mm); (<b>c</b>) Intermediate ceramic layer (thickness 0.3 ÷ 1.0 mm); (<b>d</b>) Top glass frit between intermediate and top ceramic layer (thickness 0.04 ÷ 0.06 mm); (<b>e</b>) Top ceramic layer (thickness 1.5 ÷ 2.0 mm).</p> Full article ">Figure 3
<p>Electrical scheme of Wheatstone bridges.</p> Full article ">Figure 4
<p>Wheatstone bridge.</p> Full article ">Figure 5
<p>Wheatstone bridges: (<b>a</b>) <span class="html-italic">PL</span> bridge; (<b>b</b>) 3<span class="html-italic">D</span> bridge.</p> Full article ">Figure 6
<p>Local reference system for resistors.</p> Full article ">Figure 7
<p>Global reference system for the sensor.</p> Full article ">Figure 8
<p>f.e.m. of the sensor embedded in concrete: (<b>a</b>) Concrete mesh; (<b>b</b>) Ceramic layer D mesh; (<b>c</b>) Glass-frit mesh; (<b>d</b>) Ceramic layer B mesh.</p> Full article ">Figure 9
<p>Mesh refinement: (<b>a</b>) 5352 nodes; (<b>b</b>) 7039 nodes; (<b>c</b>) 10,492 nodes.</p> Full article ">Figure 10
<p>Vertical stresses in the glass-frit layer: (<b>a</b>) Just after load; (<b>b</b>) Full viscosity developed.</p> Full article ">Figure 11
<p>Vertical stress [MPa] in 3<span class="html-italic">D</span> bridge zone in function of time (creep effect).</p> Full article ">Figure 12
<p>Planar and 3<span class="html-italic">D</span> bridge output is in the function of confinement.</p> Full article ">
17 pages, 5451 KiB  
Article
Pressure-Sensitive Capability of AgNPs Self-Sensing Cementitious Sensors
by Haoran Zhu and Min Sun
Sensors 2023, 23(24), 9629; https://doi.org/10.3390/s23249629 - 5 Dec 2023
Cited by 1 | Viewed by 1125
Abstract
Intelligent monitoring approaches for long-term, real-time digitalization in structural health monitoring (SHM) are currently attracting significant interest. Among these, self-sensing cementitious composites stand out due to their easy preparation, cost-effectiveness, and excellent compatibility with concrete structures. However, the current research faces challenges, such [...] Read more.
Intelligent monitoring approaches for long-term, real-time digitalization in structural health monitoring (SHM) are currently attracting significant interest. Among these, self-sensing cementitious composites stand out due to their easy preparation, cost-effectiveness, and excellent compatibility with concrete structures. However, the current research faces challenges, such as excessive conductive filler, difficulties in filler dispersion, and insufficient stress sensitivity and instability. This study presents a novel approach to these challenges by fabricating self-sensing cementitious sensors using silver nanoparticles (AgNPs), a new type of conductive filler. The percolation threshold of AgNPs in these materials was determined to be 0.0066 wt%, marking a reduction of approximately 90% compared to traditional conductive fillers. Moreover, the absorbance test with a UV spectrophotometer showed that AgNPs were well dispersed in an aqueous solution, which is beneficial for the construction of conductive pathways. Through various cyclic loading tests, it was observed that the self-sensing cementitious sensors with AgNPs exhibited robust pressure-sensitive stability. Additionally, their stress sensitivity reached 11.736, a value significantly surpassing that of conventional fillers. Regarding the conductive mechanism, when encountering the intricate environment within the cementitious material, AgNPs can establish numerous conductive pathways, ensuring a stable response to stress due to their ample quantity. This study provides a significant contribution to addressing the existing challenges in self-sensing cementitious materials and offers a novel reference for further research in this domain. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Aqueous solution of AgNPs.</p> Full article ">Figure 2
<p>Scanning electron micrographs of AgNPs.</p> Full article ">Figure 3
<p>Flowchart of specimen preparation.</p> Full article ">Figure 4
<p>Photographs of specimen appearance.</p> Full article ">Figure 5
<p>Flowchart of test loading for specimens.</p> Full article ">Figure 6
<p>Absorbance data of AgNP aqueous solution at different concentrations.</p> Full article ">Figure 7
<p>Comparison of absorbance data of AgNPs and CNTs.</p> Full article ">Figure 8
<p>Resistance values and their variation with time. (<b>a</b>) Changes in initial resistance value vs. drying time. (<b>b</b>) Polarization curves of G3 at different drying times.</p> Full article ">Figure 9
<p>Correspondence between specimen resistance and doping amount of AgNPs.</p> Full article ">Figure 10
<p>Cyclic loading test results of AgNP specimens. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5.</p> Full article ">Figure 10 Cont.
<p>Cyclic loading test results of AgNP specimens. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5.</p> Full article ">Figure 11
<p>Variation in ΔFCR in AgNP specimens under loading.</p> Full article ">Figure 12
<p>Results of the linear fitting of the “stress-resistivity” curve. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5. (<b>f</b>) Adj.R<sup>2</sup> of five groups.</p> Full article ">Figure 12 Cont.
<p>Results of the linear fitting of the “stress-resistivity” curve. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5. (<b>f</b>) Adj.R<sup>2</sup> of five groups.</p> Full article ">Figure 13
<p>Variable amplitude loading test results for AgNP specimens. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5.</p> Full article ">Figure 13 Cont.
<p>Variable amplitude loading test results for AgNP specimens. (<b>a</b>) G1. (<b>b</b>) G2. (<b>c</b>) G3. (<b>d</b>) G4. (<b>e</b>) G5.</p> Full article ">Figure 14
<p>Variation in SS in AgNP specimens during variable amplitude loading tests.</p> Full article ">Figure 15
<p>ΔFCR variation in AgNP specimens in variable amplitude loading tests.</p> Full article ">Figure 16
<p>Comparison of dispersion effect of fillers, where (<b>a</b>) is AgNP and (<b>b</b>) is P-CNT.</p> Full article ">Figure 17
<p>Diagram of the effect of cracks on the tunneling effect pathway.</p> Full article ">
17 pages, 7092 KiB  
Article
Analysis of FRP-Strengthened Reinforced Concrete Beams Using Electromechanical Impedance Technique and Digital Image Correlation System
by Ricardo Perera, María Consuelo Huerta, Marta Baena and Cristina Barris
Sensors 2023, 23(21), 8933; https://doi.org/10.3390/s23218933 - 2 Nov 2023
Cited by 7 | Viewed by 1378
Abstract
Fiber-reinforced polymer (FRP) strengthening systems have been considered an effective technique to retrofit concrete structures, and their use nowadays is more and more extensive. Externally bonded reinforcement (EBR) and near-surface mounted (NSM) technologies are the two most widely recognized and applied FRP strengthening [...] Read more.
Fiber-reinforced polymer (FRP) strengthening systems have been considered an effective technique to retrofit concrete structures, and their use nowadays is more and more extensive. Externally bonded reinforcement (EBR) and near-surface mounted (NSM) technologies are the two most widely recognized and applied FRP strengthening methods for enhancing structural performance worldwide. However, one of the main disadvantages of both approaches is a possible brittle failure mode provided by a sudden debonding of the FRP. Therefore, methodologies able to monitor the long-term efficiency of this kind of strengthening constitute a challenge to be overcome. In this work, two reinforced concrete (RC) specimens strengthened with FRP and subjected to increasing load tests were monitored. One specimen was strengthened using the EBR method, while for the other, the NSM technique was used. The multiple cracks emanating in both specimens in the static tests, as possible origins of a future debonding failure, were monitored using a piezoelectric (PZT)-transducer-based electromechanical impedance (EMI) technique and a digital image correlation (DIC) system. Clustering approaches based on impedance measurements of the healthy and damaged states of the specimens allowed us to suspect the occurrence of cracks and their growth. The strain profiles captured in the images of the DIC system allowed us to depict surface hair-line cracks and their propagation. The combined implementation of the two techniques to look for correlations during incremental bending tests was addressed in this study as a means of improving the prediction of early cracks and potentially anticipating the complete failure of the strengthened specimens. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Specimen cross-section.</p> Full article ">Figure 2
<p>PZTs’ set-up for Specimen 1 (dimensions in mm): (<b>a</b>) Detailed scheme; (<b>b</b>) Photo.</p> Full article ">Figure 3
<p>PZTs’ set-up for Specimen 2 (dimensions in mm).</p> Full article ">Figure 4
<p>DIC system set-up.</p> Full article ">Figure 5
<p>FRP debonding—Specimen 1.</p> Full article ">Figure 6
<p>Hierarchical trees—Specimen 1: (<b>a</b>) PZT1; (<b>b</b>) PZT2; (<b>c</b>) PZT3; (<b>d</b>) PZT4; (<b>e</b>) PZT6; (<b>f</b>) PZT7; (<b>g</b>) PZT8; (<b>h</b>) PZT9; and (<b>i</b>) PZT10.</p> Full article ">Figure 6 Cont.
<p>Hierarchical trees—Specimen 1: (<b>a</b>) PZT1; (<b>b</b>) PZT2; (<b>c</b>) PZT3; (<b>d</b>) PZT4; (<b>e</b>) PZT6; (<b>f</b>) PZT7; (<b>g</b>) PZT8; (<b>h</b>) PZT9; and (<b>i</b>) PZT10.</p> Full article ">Figure 7
<p>Evolution of the impedance signal for PZT6 sensor.</p> Full article ">Figure 8
<p>Strain map—Specimen 1: (<b>a</b>) test code 4; (<b>b</b>) test code 5; (<b>c</b>) test code 6; (<b>d</b>) test code 7; (<b>e</b>) test code 8; (<b>f</b>) test code 9; (<b>g</b>) test code 10; and (<b>h</b>) test code 11.</p> Full article ">Figure 9
<p>Failure mode—Specimen 1: (<b>a</b>) DIC image; (<b>b</b>) frontal photo; and (<b>c</b>) detail of the failure.</p> Full article ">Figure 10
<p>Hierarchical trees—Specimen 2: (<b>a</b>) PZT1; (<b>b</b>) PZT2; (<b>c</b>) PZT3; (<b>d</b>) PZT4; (<b>e</b>) PZT6; (<b>f</b>) PZT7; (<b>g</b>) PZT8; and (<b>h</b>) PZT9.</p> Full article ">Figure 11
<p>Strain map—Specimen 2: (<b>a</b>) Test 3; (<b>b</b>) Test 4; (<b>c</b>) Test 5; (<b>d</b>) Test 6; (<b>e</b>) Test 7; (<b>f</b>) Test 8; (<b>g</b>) Test 9; and (<b>h</b>) Test 10.</p> Full article ">Figure 12
<p>Concrete strain distribution.</p> Full article ">
19 pages, 10660 KiB  
Article
Corrosion Assessment in Reinforced Concrete Structures by Means of Embedded Sensors and Multivariate Analysis—Part 1: Laboratory Validation
by José Enrique Ramón-Zamora, Josep Ramon Lliso-Ferrando, Ana Martínez-Ibernón and José Manuel Gandía-Romero
Sensors 2023, 23(21), 8869; https://doi.org/10.3390/s23218869 - 31 Oct 2023
Cited by 2 | Viewed by 1507
Abstract
Reinforced Concrete Structures (RCS) are a fundamental part of a country’s civil infrastructure. However, RCSs are often affected by rebar corrosion, which poses a major problem because it reduces their service life. The traditionally used inspection and management methods applied to RCSs are [...] Read more.
Reinforced Concrete Structures (RCS) are a fundamental part of a country’s civil infrastructure. However, RCSs are often affected by rebar corrosion, which poses a major problem because it reduces their service life. The traditionally used inspection and management methods applied to RCSs are poorly operative. Structural Health Monitoring and Management (SHMM) by means of embedded sensors to analyse corrosion in RCSs is an emerging alternative, but one that still involves different challenges. Examples of SHMM include INESSCOM (Integrated Sensor Network for Smart Corrosion Monitoring), a tool that has already been implemented in different real-life cases. Nevertheless, work continues to upgrade it. To do so, the authors of this work consider implementing a new measurement procedure to identify the triggering agent of the corrosion process by analysing the double-layer capacitance of the sensors’ responses. This study was carried out on reinforced concrete specimens exposed for 18 months to different atmospheres. The results demonstrate the proposed measurement protocol and the multivariate analysis can differentiate the factor that triggers corrosion (chlorides or carbonation), even when the corrosion kinetics are similar. Data were validated by principal component analysis (PCA) and by the visual inspection of samples and rebars at the end of the study. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Structural Health Monitoring of Reinforced Concrete Structures, different approaches: advantages and disadvantages.</p> Full article ">Figure 2
<p>Test specimen description.</p> Full article ">Figure 3
<p>Corrosion potential (<math display="inline"><semantics> <mrow> <msub> <mrow> <mi>E</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> </mrow> </msub> </mrow> </semantics></math>) values. Interval plot, which shows the mean and the 95% confidence interval bars of each group.</p> Full article ">Figure 4
<p>Corrosion rate (<math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> </mrow> </msub> </mrow> </semantics></math>) values (<math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>L</mi> <mi>P</mi> <mi>R</mi> </mrow> </msub> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>P</mi> <mi>S</mi> <mi>V</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>T</mi> <mi>E</mi> </mrow> </msub> </mrow> </semantics></math>). Interval plot, which shows the mean and the 95% confidence interval bars of each group.</p> Full article ">Figure 5
<p>Regression of: (<b>a</b>) <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>L</mi> <mi>P</mi> <mi>R</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>T</mi> <mi>E</mi> </mrow> </msub> </mrow> </semantics></math> and (<b>b</b>) <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>P</mi> <mi>S</mi> <mi>V</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>T</mi> <mi>E</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 6
<p>Double-layer capacitance (<math display="inline"><semantics> <mrow> <msub> <mrow> <mi>C</mi> </mrow> <mrow> <mi>D</mi> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math>) values. Interval plot, which shows the mean and the 95% confidence interval bars of each group.</p> Full article ">Figure 7
<p>Specimens’ state at the age of 540 days (groups A, B and C).</p> Full article ">Figure 8
<p>Screen plot of the extracted principal components. The fraction of total variance represented by each principal component is also noted.</p> Full article ">Figure 9
<p>Plot of the samples on the new axes PC1-PC2 system. Samples clustering classification based on K-mean results. Respective 95% confidence ellipses are also displayed.</p> Full article ">Figure 10
<p>(<b>a</b>) Standardised loadings of the original variables for PC1 and PC2 (<b>b</b>), the position of which has the highest loadings on one of the polarisation curves (<math display="inline"><semantics> <mrow> <mo>∆</mo> <mi>I</mi> </mrow> </semantics></math> − <math display="inline"><semantics> <mrow> <mo>∆</mo> <mi>E</mi> </mrow> </semantics></math>).</p> Full article ">Figure 11
<p>Spontaneous clustering of samples when <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> <mo>−</mo> <mi>P</mi> <mi>S</mi> <mi>V</mi> </mrow> </msub> </mrow> </semantics></math> vs. <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>C</mi> </mrow> <mrow> <mi>D</mi> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math> are plotted. Respective 95% confidence ellipse and a hypothetical position of two axes equivalent to the PC1-PC2 axes obtained by PCA (dashed line) are displayed. In addition, the corrosion risk intervals are plotted according to the ranges established for <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>i</mi> </mrow> <mrow> <mi>C</mi> <mi>O</mi> <mi>R</mi> <mi>R</mi> </mrow> </msub> </mrow> </semantics></math> by Standard UNE 112072:2011 [<a href="#B83-sensors-23-08869" class="html-bibr">83</a>].</p> Full article ">
12 pages, 4846 KiB  
Communication
Novel Weigh-in-Motion Pavement Sensor Based on Self-Sensing Nanocomposites for Vehicle Load Identification: Development, Performance Testing, and Validation
by Ming Liang, Yunfeng Zhang, Yuepeng Jiao, Jianjiang Wang, Linping Su and Zhanyong Yao
Sensors 2023, 23(10), 4758; https://doi.org/10.3390/s23104758 - 15 May 2023
Cited by 2 | Viewed by 1938
Abstract
The development of the transportation industry has led to an increasing number of overloaded vehicles, which reduces the service life of asphalt pavements. Currently, the traditional vehicle weighing method not only involves heavy equipment but also has a low weighing efficiency. To deal [...] Read more.
The development of the transportation industry has led to an increasing number of overloaded vehicles, which reduces the service life of asphalt pavements. Currently, the traditional vehicle weighing method not only involves heavy equipment but also has a low weighing efficiency. To deal with the defects in the existing vehicle weighing system, this paper developed a road-embedded piezoresistive sensor based on self-sensing nanocomposites. The sensor developed in this paper adopts an integrated casting and encapsulation technology, in which an epoxy resin/MWCNT nanocomposite is used for the functional phase, and an epoxy resin/anhydride curing system is used for the high-temperature resistant encapsulation phase. The compressive stress-resistance response characteristics of the sensor were investigated by calibration experiments with an indoor universal testing machine. In addition, the sensors were embedded in the compacted asphalt concrete to validate the applicability to the harsh environment and back-calculate the dynamic vehicle loads on the rutting slab. The results show that the response relationship between the sensor resistance signal and the load is in accordance with the GaussAmp formula. The developed sensor not only survives effectively in asphalt concrete but also enables dynamic weighing of the vehicle loads. Consequently, this study provides a new pathway to develop high-performance weigh-in-motion pavement sensors. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic diagram of integration mold for sensing material.</p> Full article ">Figure 2
<p>Piezoresistive sensor based on composite smart materials.</p> Full article ">Figure 3
<p>Calibration test via universal testing machine.</p> Full article ">Figure 4
<p>Fitting diagram of compressive stress–resistance response at different loading rates. (<b>a</b>) 20 N/s, (<b>b</b>) 80 N/s, (<b>c</b>) 140 N/s, (<b>d</b>) 200 N/s, and (<b>e</b>) 260 N/s.</p> Full article ">Figure 5
<p>Epoxy base for positioning.</p> Full article ">Figure 6
<p>The embedment of piezoresistive sensors and formation of rutting slab.</p> Full article ">Figure 7
<p>Compressive stress response of sensors in asphalt concrete.</p> Full article ">Figure 8
<p>Rutting experiment for axle load sensor.</p> Full article ">Figure 9
<p>Compressive stress–resistance response during rut experiment.</p> Full article ">
14 pages, 6632 KiB  
Article
Development of Deep Belief Network for Tool Faults Recognition
by Archana P. Kale, Revati M. Wahul, Abhishek D. Patange, Rohan Soman and Wieslaw Ostachowicz
Sensors 2023, 23(4), 1872; https://doi.org/10.3390/s23041872 - 7 Feb 2023
Cited by 21 | Viewed by 2230
Abstract
The controlled interaction of work material and cutting tool is responsible for the precise outcome of machining activity. Any deviation in cutting parameters such as speed, feed, and depth of cut causes a disturbance to the machining. This leads to the deterioration of [...] Read more.
The controlled interaction of work material and cutting tool is responsible for the precise outcome of machining activity. Any deviation in cutting parameters such as speed, feed, and depth of cut causes a disturbance to the machining. This leads to the deterioration of a cutting edge and unfinished work material. Recognition and description of tool failure are essential and must be addressed using intelligent techniques. Deep learning is an efficient method that assists in dealing with a large amount of dynamic data. The manufacturing industry generates momentous information every day and has enormous scope for data analysis. Most intelligent systems have been applied toward the prediction of tool conditions; however, they must be explored for descriptive analytics for on-board pattern recognition. In an attempt to recognize the variation in milling operation leading to tool faults, the development of a Deep Belief Network (DBN) is presented. The network intends to classify in total six tool conditions (one healthy and five faulty) through image-based vibration signals acquired in real time. The model was designed, trained, tested, and validated through datasets collected considering diverse input parameters. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Step-wise organization of the current investigation.</p> Full article ">Figure 2
<p>The schematic of the experimentation arrangement.</p> Full article ">Figure 3
<p>STFT Spectrogram representing various tool conditions.</p> Full article ">Figure 4
<p>Mechanism of Deep Belief Network and Deep Boltzmann Machine.</p> Full article ">Figure 5
<p>Architecture of Deep Belief Network.</p> Full article ">Figure 6
<p>Confusion matrix considering cross-validation mode (ADF: All defect free (healthy), FL: Flank wear, NS: Nose wear, NT: Notch, CT: Crater, AD: All defective).</p> Full article ">Figure 7
<p>Confusion matrix considering training dataset.</p> Full article ">Figure 8
<p>Confusion matrix considering test dataset.</p> Full article ">Figure 9
<p>Confusion matrix considering blind dataset.</p> Full article ">
14 pages, 1239 KiB  
Article
Electromechanical Reciprocity Applied to the Sensing Properties of Guided Elastic Wave Transducers
by Bernd Köhler, Lars Schubert, Martin Barth and Kazuyuki Nakahata
Sensors 2023, 23(1), 150; https://doi.org/10.3390/s23010150 - 23 Dec 2022
Cited by 1 | Viewed by 1372
Abstract
Guided elastic wave (GEW) transducers for structural health monitoring (SHM) can act as transmitters (senders) and receivers (sensors). Their performance in both cases depends on the structure to which they are coupled. Therefore, they must be characterized as system transducer- structure. The characterization [...] Read more.
Guided elastic wave (GEW) transducers for structural health monitoring (SHM) can act as transmitters (senders) and receivers (sensors). Their performance in both cases depends on the structure to which they are coupled. Therefore, they must be characterized as system transducer- structure. The characterization of the transducer-structure as transmitter using a Scanning Laser Doppler Vibrometer (SLDV) is straightforward, whereas its characterization as receiver is non-trivial. We propose to exploit electromechanical reciprocity, which is an identity between the transfer functions of electrical-to-mechanical and mechanical-to-electrical conversions. For this purpose, the well-known electromechanical reciprocity theorem was adapted to the following situation: The two reciprocal states are “electrical excitation and detection of the surface velocity at point P” and “mechanical excitation at P and measurement of the electrical quantities”. According to the derived formulas, the quantities on the mechanical and electrical sides must be chosen appropriately to ensure reciprocity as well as that the corresponding transfer functions are equal. We demonstrate the reciprocity with experimental data for correctly chosen transfer functions and show the deviation in reciprocity for a different choice. Furthermore, we propose further applications of electromechanical reciprocity. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Piezoelectric probe coupled to a solid test specimen; the integration volume V includes the test specimen and the probe and can be of any shape. V has the boundary S indicated by the dashed line. The part of the boundary crossing the electrical coaxial cable is denoted as S<sub>cable</sub>. The point P at the surface has coordinates <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">r</mi> <mn>0</mn> </msub> </mrow> </semantics></math> where in state “B” an external point force acts to the surface and the surface velocity is measured in state “A”.</p> Full article ">Figure 2
<p>Schematic of experimental setup. (<b>Left</b>): the transducer acts as the sender and the out-of-plane velocity is measured. (<b>Right</b>): the transducer acts as a receiver, the plate is excited by a point force, and the voltage response under open-circuit conditions is measured. The dimensions of the plate are 1000 mm × 1000 mm × 2 mm.</p> Full article ">Figure 3
<p>Electric current into the transducer when it acts as sender. The current <math display="inline"><semantics> <mrow> <mi>I</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </semantics></math> was low pass filtered with a cut-off frequency of 2 MHz.</p> Full article ">Figure 4
<p>Comparison of the velocity signals. (<b>Top row</b>): velocity measured using LDV as response to the current signal <math display="inline"><semantics> <mrow> <msup> <mi>I</mi> <mrow> <mi>s</mi> <mi>e</mi> <mi>n</mi> <mi>d</mi> </mrow> </msup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </semantics></math>. <b>(Middle row</b>): quantity <math display="inline"><semantics> <mrow> <msup> <mi>v</mi> <mrow> <mi>c</mi> <mi>a</mi> <mi>l</mi> <mi>c</mi> </mrow> </msup> <mo>=</mo> <mi>I</mi> <mo>∗</mo> <mi>t</mi> <msup> <mi>r</mi> <mrow> <mi>r</mi> <mi>e</mi> <mi>c</mi> </mrow> </msup> <mo>~</mo> <mi>I</mi> <mo>∗</mo> <mi>U</mi> <mo>′</mo> </mrow> </semantics></math> and (<b>Bottom row</b>): the quantity <math display="inline"><semantics> <mrow> <msup> <mi>v</mi> <mrow> <mi>c</mi> <mi>a</mi> <mi>l</mi> <mi>c</mi> <mo>_</mo> <mi>I</mi> <mi>I</mi> </mrow> </msup> <mo>=</mo> <msup> <mi>U</mi> <mrow> <mi>s</mi> <mi>e</mi> <mi>n</mi> <mi>d</mi> </mrow> </msup> <mo>∗</mo> <mi>t</mi> <msup> <mi>r</mi> <mrow> <mi>r</mi> <mi>e</mi> <mi>c</mi> </mrow> </msup> <mo>~</mo> <mi>U</mi> <mo>′</mo> </mrow> </semantics></math> which must be equal to the measured <math display="inline"><semantics> <mrow> <mi>v</mi> </mrow> </semantics></math> when the “presumed” reciprocity would be valid. The arrows indicate positions where <math display="inline"><semantics> <mrow> <msup> <mi>v</mi> <mrow> <mi>c</mi> <mi>a</mi> <mi>l</mi> <mi>c</mi> <mo>_</mo> <mi>I</mi> <mi>I</mi> </mrow> </msup> </mrow> </semantics></math> significantly deviates from <math display="inline"><semantics> <mi>v</mi> </semantics></math>.</p> Full article ">Figure 5
<p>Comparison of the measured surface normal velocity <math display="inline"><semantics> <mi>v</mi> </semantics></math> with transducer as sender and the given (non <math display="inline"><semantics> <mi>δ</mi> </semantics></math>-pulse like) current excitation with <math display="inline"><semantics> <mrow> <msup> <mi>v</mi> <mrow> <mi>c</mi> <mi>a</mi> <mi>l</mi> <mi>c</mi> </mrow> </msup> </mrow> </semantics></math> calculated according to (16) from the data of the transducer as receiver. Both quantities are shown at arbitrary scales.</p> Full article ">
28 pages, 11050 KiB  
Article
Performance Assessment for a Guided Wave-Based SHM System Applied to a Stiffened Composite Structure
by Inka Mueller, Vittorio Memmolo, Kilian Tschöke, Maria Moix-Bonet, Kathrin Möllenhoff, Mikhail Golub, Ramanan Sridaran Venkat, Yevgeniya Lugovtsova, Artem Eremin and Jochen Moll
Sensors 2022, 22(19), 7529; https://doi.org/10.3390/s22197529 - 4 Oct 2022
Cited by 11 | Viewed by 2319
Abstract
To assess the ability of structural health monitoring (SHM) systems, a variety of prerequisites and contributing factors have to be taken into account. Within this publication, this variety is analyzed for actively introduced guided wave-based SHM systems. For these systems, it is not [...] Read more.
To assess the ability of structural health monitoring (SHM) systems, a variety of prerequisites and contributing factors have to be taken into account. Within this publication, this variety is analyzed for actively introduced guided wave-based SHM systems. For these systems, it is not possible to analyze their performance without taking into account their structure and their applied system parameters. Therefore, interdependencies of performance assessment are displayed in an SHM pyramid based on the structure and its monitoring requirements. Factors influencing the quality, capability and reliability of the monitoring system are given and put into relation with state-of-the-art performance analysis in a non-destructive evaluation. While some aspects are similar and can be treated in similar ways, others, such as location, environmental condition and structural dependency, demand novel solutions. Using an open-access data set from the Open Guided Waves platform, a detailed method description and analysis of path-based performance assessment is presented.The adopted approach clearly begs the question about the decision framework, as the threshold affects the reliability of the system. In addition, the findings show the effect of the propagation path according to the damage position. Indeed, the distance of damage directly affects the system performance. Otherwise, the propagation direction does not alter the potentiality of the detection approach despite the anisotropy of composites. Nonetheless, the finite waveguide makes it necessary to look at the whole paths, as singular phenomena associated with the reflections may appear. Numerical investigation helps to clarify the centrality of wave mechanics and the necessity to take sensor position into account as an influencing factor. Starting from the findings achieved, all the issues are discussed, and potential future steps are outlined. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Pyramid of Prerequisites for Performance Assessment of SHM Systems.</p> Full article ">Figure 2
<p>Probability of detection curve versus flaw size, calculated using the hit/miss procedure with the function model and estimation of coefficients.</p> Full article ">Figure 3
<p><math display="inline"><semantics> <mover accent="true"> <mi>a</mi> <mo>^</mo> </mover> </semantics></math> vs. <span class="html-italic">a</span> procedure with the regression model and the Gaussian distribution around the predicted value (illustrative data).</p> Full article ">Figure 4
<p>Probability of detection curve versus flaw size, calculated using the <math display="inline"><semantics> <mover accent="true"> <mi>a</mi> <mo>^</mo> </mover> </semantics></math> vs <span class="html-italic">a</span> procedure with the definition of Berens and integral formulation (illustrative data).</p> Full article ">Figure 5
<p>Confidence bounds of the predicted response (illustrative data).</p> Full article ">Figure 6
<p>Specimen geometry with the three damage positions D1–D3 relative to the transducer locations T1–T12 ([<a href="#B4-sensors-22-07529" class="html-bibr">4</a>]). Considering path T3–T9, <math display="inline"><semantics> <msub> <mi>d</mi> <mn>3</mn> </msub> </semantics></math> represents the distance from damage D3, while <math display="inline"><semantics> <msub> <mi>θ</mi> <mn>3</mn> </msub> </semantics></math> is the angle of the transducer pair path against the damage orientation, which is at 45<math display="inline"><semantics> <msup> <mrow/> <mo>∘</mo> </msup> </semantics></math> respect to x-axis.</p> Full article ">Figure 7
<p>Illustration of the experimental setup arranged in a climatic chamber for the measurement campaign. The plate is instrumented with several transducers and damaged by artificial defect.</p> Full article ">Figure 8
<p><math display="inline"><semantics> <mover accent="true"> <mi>a</mi> <mo>^</mo> </mover> </semantics></math> vs. <span class="html-italic">a</span> regression analysis. Data samples related to damage D1 and paths T3–T9 and T4–T7. The corresponding linear trend is established using <math display="inline"><semantics> <mrow> <mi>l</mi> <mi>i</mi> <mi>n</mi> <mo>−</mo> <mi>l</mi> <mi>i</mi> <mi>n</mi> </mrow> </semantics></math> (<b>a</b>) and <math display="inline"><semantics> <mrow> <mi>l</mi> <mi>o</mi> <mi>g</mi> <mo>−</mo> <mi>l</mi> <mi>o</mi> <mi>g</mi> </mrow> </semantics></math> (<b>b</b>) models.</p> Full article ">Figure 9
<p>Noise and signal response samples along the path T3–T9 and damage D1. The regression and the signal response distribution are established using <math display="inline"><semantics> <mrow> <mi>l</mi> <mi>o</mi> <mi>g</mi> <mo>−</mo> <mi>l</mi> <mi>o</mi> <mi>g</mi> </mrow> </semantics></math> model. The <span class="html-italic">a</span>-value corresponding to noise samples is zero (no-damage) and is moved along <span class="html-italic">x</span>-axis for a better visualization.</p> Full article ">Figure 10
<p>POD versus flaw size related to damage D1 and paths T3–T9 and T4–T7. Regression model used to estimate <math display="inline"><semantics> <mover accent="true"> <mi>a</mi> <mo>^</mo> </mover> </semantics></math> is according to <a href="#sensors-22-07529-f008" class="html-fig">Figure 8</a>b.</p> Full article ">Figure 11
<p>Histogram for damage index values for different paths, (<b>a</b>) T1–T7, (<b>b</b>) T7–T11, and (<b>c</b>) all paths, as well as those that do not cross the stringer. (<b>d</b>) Empirical CDF of damage index values given for all paths.</p> Full article ">Figure 12
<p>Resulting threshold values <math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mi>d</mi> <mi>e</mi> <mi>c</mi> </mrow> </msub> </semantics></math> based on empirical cumulative distribution of damage index values given for the cumulative empirical cdf of all paths and cdfs of exemplary paths.The fields are colored according to their value; small values are colored green, and high values are colored red.</p> Full article ">Figure 13
<p>POD curves for all non-horizontal paths, which all cross the stringer over the distance. The right figure shows a detail of the left zooming in to <span class="html-italic">a</span>-values from 0 mm<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math> to 500 mm<math display="inline"><semantics> <msup> <mrow/> <mn>2</mn> </msup> </semantics></math>.</p> Full article ">Figure 14
<p><math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> versus path absolute distance from the flaw as a 2D representation of <a href="#sensors-22-07529-f013" class="html-fig">Figure 13</a>.</p> Full article ">Figure 15
<p><math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> values versus interrogation path. The <math display="inline"><semantics> <mrow> <mi>POD</mi> <mo>(</mo> <mi>a</mi> <mo>)</mo> </mrow> </semantics></math> is predicted along all vertical paths. Damage D1 (<b>a</b>), D2 (<b>b</b>), and D3 (<b>c</b>).</p> Full article ">Figure 16
<p><math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> versus path distance from the flaw. The <math display="inline"><semantics> <mrow> <mi>POD</mi> <mo>(</mo> <mi>a</mi> <mo>)</mo> </mrow> </semantics></math> is predicted along all paths crossing the stringer.</p> Full article ">Figure 17
<p>Interpolating surface of <math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> values versus path-damage distance and incidence angle (<math display="inline"><semantics> <mi>θ</mi> </semantics></math>) from the flaw. The <math display="inline"><semantics> <mrow> <mi>POD</mi> <mo>(</mo> <mi>a</mi> <mo>)</mo> </mrow> </semantics></math> is predicted along all through the paths crossing the stringer.</p> Full article ">Figure 18
<p>Visualisation of the <math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> values along different paths. A circle ∘ marks the damage position; the crosses × mark transducers. The evaluation was done for <math display="inline"><semantics> <mrow> <mi>D</mi> <msub> <mi>I</mi> <mrow> <mi>E</mi> <mi>n</mi> <mi>e</mi> <mi>r</mi> <mi>g</mi> <mi>y</mi> </mrow> </msub> </mrow> </semantics></math> at <math display="inline"><semantics> <mrow> <mi>f</mi> <mo>=</mo> <mn>40</mn> <mspace width="3.33333pt"/> </mrow> </semantics></math> kHz and <math display="inline"><semantics> <mrow> <msub> <mi>a</mi> <mrow> <mi>d</mi> <mi>e</mi> <mi>c</mi> </mrow> </msub> <mo>=</mo> <mn>0.01</mn> </mrow> </semantics></math>.</p> Full article ">Figure 18 Cont.
<p>Visualisation of the <math display="inline"><semantics> <msub> <mi>a</mi> <mrow> <mn>90</mn> <mo>|</mo> <mn>95</mn> </mrow> </msub> </semantics></math> values along different paths. A circle ∘ marks the damage position; the crosses × mark transducers. The evaluation was done for <math display="inline"><semantics> <mrow> <mi>D</mi> <msub> <mi>I</mi> <mrow> <mi>E</mi> <mi>n</mi> <mi>e</mi> <mi>r</mi> <mi>g</mi> <mi>y</mi> </mrow> </msub> </mrow> </semantics></math> at <math display="inline"><semantics> <mrow> <mi>f</mi> <mo>=</mo> <mn>40</mn> <mspace width="3.33333pt"/> </mrow> </semantics></math> kHz and <math display="inline"><semantics> <mrow> <msub> <mi>a</mi> <mrow> <mi>d</mi> <mi>e</mi> <mi>c</mi> </mrow> </msub> <mo>=</mo> <mn>0.01</mn> </mrow> </semantics></math>.</p> Full article ">Figure 19
<p>Exemplary representation of the wave propagation in the test specimen. The wave field of the differential signal (undamaged–damaged) is shown for a damage of size 671 mm<sup>2</sup> at position D1 and the actuator-sensor path T4–T7. The time steps <math display="inline"><semantics> <mrow> <mn>150</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s, <math display="inline"><semantics> <mrow> <mn>250</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s and <math display="inline"><semantics> <mrow> <mn>350</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s were selected.</p> Full article ">Figure 20
<p>Exemplary representation of the wave propagation in the test specimen. The wave field of the differential signal (undamaged–damaged) is shown for a damage of size 671 mm<sup>2</sup> at position D1 and the actuator-sensor path T2–T11. The time steps <math display="inline"><semantics> <mrow> <mn>150</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s, <math display="inline"><semantics> <mrow> <mn>250</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s and <math display="inline"><semantics> <mrow> <mn>350</mn> <mo> </mo> <mi mathvariant="sans-serif">μ</mi> </mrow> </semantics></math>s were selected.</p> Full article ">Figure 21
<p>Exemplary differential signals of the OGW data set for a damage of size 671 mm<sup>2</sup> at position D1 and the mentioned actuator-sensor paths.</p> Full article ">
23 pages, 7981 KiB  
Article
Sub-Surface Defect Depth Approximation in Cold Infrared Thermography
by Siavash Doshvarpassand and Xiangyu Wang
Sensors 2022, 22(18), 7098; https://doi.org/10.3390/s22187098 - 19 Sep 2022
Cited by 4 | Viewed by 2648
Abstract
Detection and characterisation of hidden corrosion are considered challenging yet crucial activities in many sensitive industrial plants where preventing the loss of containment or structural reliability are paramount. In the last two decades, infrared (IR) thermography has proved to be a reliable means [...] Read more.
Detection and characterisation of hidden corrosion are considered challenging yet crucial activities in many sensitive industrial plants where preventing the loss of containment or structural reliability are paramount. In the last two decades, infrared (IR) thermography has proved to be a reliable means for inspection of corrosion or other sub-surface anomalies in low to mid thickness metallic mediums. The foundation of using IR thermography for defect detection and characterisation is based on active thermography. In this method of inspection, an external excitation source is deployed for the purpose of stimulating thermal evolutions inside objects. The presence of sub-surface defects disrupts the evolution of electromagnetic pulse inside an object. The reflection of altered pulse at the surface can be recorded through thermal camera in the form of temperature anomalies. Through authors’ previous works, cold thermography has shown that it can be a viable defect detection alternative to the most commonly used means of active thermography, known as heating. In the current work, the characterisation of defect dimensions, i.e., depth and diameter, has been explored. A simple analytical model for thermal contrast over defect is used in order to approximate the defect depth and diameter. This is achieved by comparing the similarities of the model and the experimental contrast time-series. A method of time-series similarity measurement known as dynamic time wrapping (DTW) is used to score the similarity between a pair of model and experiment time-series. The final outcome of the proposed experimental setup has revealed that there is a good potential to predict the metal loss of up to 50% in mid-thickness substrate even by deploying a less accurate nonradiometric thermal device and no advanced image processing. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic of thermal diffusion and response through a defective solid, retrieved from [<a href="#B1-sensors-22-07098" class="html-bibr">1</a>].</p> Full article ">Figure 2
<p>Comparison of the effect of defect (<b>a</b>) infinite and (<b>b</b>) finite lateral extension on heat diffusion, retrieved from [<a href="#B1-sensors-22-07098" class="html-bibr">1</a>].</p> Full article ">Figure 3
<p>(<b>a</b>) Carrier front view; (<b>b</b>) carrier side view; (<b>c</b>) carrier back view; (<b>d</b>) cooling medium spray with customised flat-fan nozzle; (<b>e</b>) experimental setup showing carrier motion over the test piece.</p> Full article ">Figure 4
<p>The test piece and the defect arrangements and dimensions.</p> Full article ">Figure 5
<p>Statistical parameters representing the uniformity and consistency of the carrier speed. (<b>a</b>) Shows the histogram of test speeds and their equivalent distributions for 12 tests and (<b>b</b>) represents Kurtosis and Skewness values calculated from each distribution.</p> Full article ">Figure 6
<p>Experimental data ingestion, preprocessing and analytic pipeline steps.</p> Full article ">Figure 7
<p>(<b>a</b>,<b>b</b>) 1D temperature and running temperature contrast defects with diameter <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> and various depths, <math display="inline"><semantics> <mrow> <mi>d</mi> </mrow> </semantics></math>, on an <math display="inline"><semantics> <mrow> <mi>L</mi> <mo>=</mo> <mn>8</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> thickness mild steel specimen subjected to a square-pulse stimulation of duration <math display="inline"><semantics> <mrow> <msub> <mi>t</mi> <mrow> <mi>s</mi> <mi>t</mi> <mi>m</mi> </mrow> </msub> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math> s. (<b>c</b>,<b>d</b>) 1D temperature and running temperature contrast of a defect with diameter <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>d</mi> <mo>=</mo> <mn>1</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> on an <math display="inline"><semantics> <mrow> <mi>L</mi> <mo>=</mo> <mn>8</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> thickness mild steel sample subjected to square-pulse stimulation of duration <math display="inline"><semantics> <mrow> <msub> <mi>t</mi> <mrow> <mi>s</mi> <mi>t</mi> <mi>m</mi> </mrow> </msub> <mo>=</mo> <mn>0.5</mn> <mo>,</mo> <mo> </mo> <mn>1</mn> <mo> </mo> <mi>and</mi> <mo> </mo> <mn>2</mn> </mrow> </semantics></math> s. The pulse absorbed power density is assumed <math display="inline"><semantics> <mrow> <mi>Q</mi> <mo>=</mo> <mn>10</mn> <mo> </mo> <mi>kW</mi> <mo>/</mo> <msup> <mi mathvariant="normal">m</mi> <mn>2</mn> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 8
<p>(<b>a</b>) Average running contrast of <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mo> </mo> <mi>mm</mi> <mo>,</mo> <mo> </mo> <mi>d</mi> <mo>=</mo> <mn>1</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> defect for various test speeds; (<b>b</b>) average running contrast of <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> defects with different depths under forward and backward regimes.</p> Full article ">Figure 9
<p>Contrast peak amplitude ratio of (<b>a</b>) various test direction and (<b>b</b>) various test speed regimes for various defect diameters and depths.</p> Full article ">Figure 10
<p>(<b>a</b>) Analytical (3D) running contrast defects for various diameter, <math display="inline"><semantics> <mrow> <mi>D</mi> </mrow> </semantics></math>, on an <math display="inline"><semantics> <mrow> <mi>L</mi> <mo>=</mo> <mn>8</mn> <mrow> <mo> </mo> <mi>mm</mi> </mrow> </mrow> </semantics></math> thickness mild steel specimen subjected to a square-pulse stimulation of duration <math display="inline"><semantics> <mrow> <msub> <mi>t</mi> <mrow> <mi>s</mi> <mi>t</mi> <mi>m</mi> </mrow> </msub> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math> s. (<b>b</b>) Analytical (3D) running contrast of a defect with diameter <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mrow> <mo> </mo> <mi>mm</mi> </mrow> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>d</mi> <mo>=</mo> <mn>1</mn> <mrow> <mo> </mo> <mi>mm</mi> </mrow> </mrow> </semantics></math> on an <math display="inline"><semantics> <mrow> <mi>L</mi> <mo>=</mo> <mn>8</mn> <mo> </mo> <mi>mm</mi> </mrow> </semantics></math> thickness mild steel sample subjected to square-pulse stimulation of duration <math display="inline"><semantics> <mrow> <msub> <mi>t</mi> <mrow> <mi>s</mi> <mi>t</mi> <mi>m</mi> </mrow> </msub> <mo>=</mo> <mn>0.5</mn> <mo>,</mo> <mo> </mo> <mn>1</mn> <mrow> <mo> </mo> <mi>and</mi> <mo> </mo> </mrow> <mn>2</mn> </mrow> </semantics></math> s. The pulse-absorbed power density is assumed <math display="inline"><semantics> <mrow> <mi>Q</mi> <mo>=</mo> <mn>10</mn> <mo> </mo> <mi>kW</mi> <mo>/</mo> <msup> <mi mathvariant="normal">m</mi> <mn>2</mn> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 11
<p>Contrast peak amplitude ratio of (<b>a</b>) 0.5 and 1 s and (<b>b</b>) 1 and 2 s pulse duration for various defect diameters and depths.</p> Full article ">Figure 12
<p>A comparison between experimental results and analytical model for defects of diameter 22 and 18 mm for various depths.</p> Full article ">Figure 13
<p>The ratio of contrast peak amplitude for experiments against model.</p> Full article ">Figure 14
<p>A comparison between experimental results, analytical model and adjusted model for defects of diameter 22 and 18 mm for various depths.</p> Full article ">Figure 15
<p>A comparison of Euclidean and DTW alignments of experimental and analytical model of running contrast curve for defect <math display="inline"><semantics> <mrow> <mi>D</mi> <mo>=</mo> <mn>22</mn> <mrow> <mo> </mo> <mi>mm</mi> </mrow> <mo>,</mo> <mo> </mo> <mi>d</mi> <mo>=</mo> <mn>1</mn> <mrow> <mo> </mo> <mi>mm</mi> </mrow> </mrow> </semantics></math>.</p> Full article ">Figure 16
<p>Defect depth and diameter prediction using DTW method for the forward and backward test regimes. Green color reflects the 10% percentile minimum range of similarity scores and change of colours towards red is equivalent to reduction of similarities (increasing similarity score) between experimental contrast and analytical model contrast time-series.</p> Full article ">
16 pages, 5805 KiB  
Article
Application of a Wireless and Contactless Ultrasonic System to Evaluate Optimal Sawcut Time for Concrete Pavements
by Homin Song, Jinyoung Hong, Young-Geun Yoon, Hajin Choi and Taekeun Oh
Sensors 2022, 22(18), 7030; https://doi.org/10.3390/s22187030 - 16 Sep 2022
Cited by 2 | Viewed by 1862
Abstract
A recently developed contactless ultrasonic testing scheme is applied to define the optimal saw-cutting time for concrete pavement. The ultrasonic system is improved using wireless data transfer for field applications, and the signal processing and data analysis are proposed to evaluate the modulus [...] Read more.
A recently developed contactless ultrasonic testing scheme is applied to define the optimal saw-cutting time for concrete pavement. The ultrasonic system is improved using wireless data transfer for field applications, and the signal processing and data analysis are proposed to evaluate the modulus of elasticity of early-age concrete. Numerical simulation of leaky Rayleigh wave in joint-half space including air and concrete is performed to demonstrate the proposed data analysis procedure. The hardware and algorithms developed for the ultrasonic system are experimentally validated with a comparison of saw-cutting procedures. In addition, conventional methods for the characterization of early-age concrete, including pin penetration and maturity methods, are applied. The results demonstrated that the developed wireless system presents identical results to the wired system, and the initiation time of leaky Rayleigh wave possibly represents 5% of raveling damage compared to the optimal saw cutting. Further data analysis implies that saw-cutting would be optimally performed at approximately 11.5 GPa elastic modulus of concrete obtained by the wireless and contactless ultrasonic system. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Wavefield snapshots generated from numerical simulations: ultrasonic wavefields at t = 300 µs for the cases of (<b>a</b>) beginning of hydration (<span class="html-italic">E</span> = 0.1 GPa) and (<b>b</b>) fully hydrated concrete (<span class="html-italic">E</span> = 25 GPa).</p> Full article ">Figure 2
<p>Derived relationship between the leaky Rayleigh wave velocity (<math display="inline"><semantics> <mrow> <msub> <mi>V</mi> <mrow> <mi>L</mi> <mi>R</mi> </mrow> </msub> <mo stretchy="false">)</mo> </mrow> </semantics></math> and elastic modulus (<span class="html-italic">E</span>). Note that the range of theoretical Rayleigh wave velocity was calculated across the range of Poisson’s ratio (0.15~0.3) and mass density (2200~2600 kg/m<sup>3</sup>).</p> Full article ">Figure 3
<p>The developed wireless and contactless ultrasonic measurement system.</p> Full article ">Figure 4
<p>A flowchart of the proposed signal processing scheme.</p> Full article ">Figure 5
<p>Test setups: (<b>a</b>) contactless ultrasonic systems (wired and wireless systems) with thermal gauges, (<b>b</b>) pin penetration test setup, and (<b>c</b>) a slab specimen for saw cuts.</p> Full article ">Figure 6
<p>B-scan images obtained from the (<b>a</b>) wired system, and (<b>b</b>) wireless system.</p> Full article ">Figure 7
<p>Test results from the (<b>a</b>) pin penetration test, and (<b>b</b>) maturity method.</p> Full article ">Figure 8
<p>Saw cut damage identification: (<b>a</b>) saw cut images cropped from photos, and (<b>b</b>) the corresponding binarized images. Note that the red solid box indicates the windowed region, and the yellow box indicates the RoI.</p> Full article ">Figure 9
<p>The obtained saw-cut areas indicating two distinct regions showing different slopes.</p> Full article ">Figure 10
<p>Elastic modulus of concrete computed using measured leaky Rayleigh wave responses. Elapsed time is presented in a colormap.</p> Full article ">Figure 11
<p>Correlation between the damage index of a saw-cut area and elastic modulus of concrete.</p> Full article ">
18 pages, 4290 KiB  
Article
Loosening Identification of Multi-Bolt Connections Based on Wavelet Transform and ResNet-50 Convolutional Neural Network
by Xiao-Xue Li, Dan Li, Wei-Xin Ren and Jun-Shu Zhang
Sensors 2022, 22(18), 6825; https://doi.org/10.3390/s22186825 - 9 Sep 2022
Cited by 22 | Viewed by 2476
Abstract
A high-strength bolt connection is the key component of large-scale steel structures. Bolt loosening and preload loss during operation can reduce the load-carrying capacity, safety, and durability of the structures. In order to detect loosening damage in multi-bolt connections of large-scale civil engineering [...] Read more.
A high-strength bolt connection is the key component of large-scale steel structures. Bolt loosening and preload loss during operation can reduce the load-carrying capacity, safety, and durability of the structures. In order to detect loosening damage in multi-bolt connections of large-scale civil engineering structures, we proposed a multi-bolt loosening identification method based on time-frequency diagrams and a convolutional neural network (CNN) using vi-bro-acoustic modulation (VAM) signals. Continuous wavelet transform was employed to obtain the time-frequency diagrams of VAM signals as the features. Afterward, the CNN model was trained to identify the multi-bolt loosening conditions from the raw time-frequency diagrams intelligently. It helps to get rid of the dependence on traditional manual selection of simplex and ineffective damage index and to eliminate the influence of operational noise of structures on the identification accuracy. A laboratory test was carried out on bolted connection specimens with four high-strength bolts of different degrees of loosening. The effects of different excitations, CNN models, and dataset sizes were investigated. We found that the ResNet-50 CNN model taking time-frequency diagrams of the hammer excited VAM signals, as the input had better performance in identifying the loosened bolts with various degrees of loosening at different positions. The results indicate that the proposed multi-bolt loosening identification method based on VAM and ResNet-50 CNN can identify bolt loosening with a reasonable accuracy, computational efficiency, and robustness. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Principle of vibro-acoustic modulation (VAM) technique.</p> Full article ">Figure 2
<p>ResNet-50 residual block.</p> Full article ">Figure 3
<p>ResNet-50 neural network configuration.</p> Full article ">Figure 4
<p>Multi-bolt loosening identification flowchart.</p> Full article ">Figure 5
<p>Experimental setup.</p> Full article ">Figure 6
<p>Time-frequency diagram obtained using different test methods excitation frequencies: (<b>a</b>) results of vibration generator test under case J; (<b>b</b>) results of hammer test under case J; (<b>c</b>) results of vibration generator test under case A; (<b>d</b>) results of hammer test under case A.</p> Full article ">Figure 7
<p>Image augmentation.</p> Full article ">Figure 8
<p>Comparison of training results between ResNet-50 and other CNN models: (<b>a</b>) training accuracy; (<b>b</b>) validation accuracy; (<b>c</b>) training loss; and (<b>d</b>) validation loss.</p> Full article ">Figure 9
<p>Comparison of different excitation forms on model performance of ResNet-50 model. (<b>a</b>) training accuracy; (<b>b</b>) validation accuracy; (<b>c</b>) training loss; and (<b>d</b>) validation loss.</p> Full article ">Figure 10
<p>Confusion matrix.</p> Full article ">
16 pages, 8411 KiB  
Article
Monitoring Osseointegration Process Using Vibration Analysis
by Shouxun Lu, Benjamin Steven Vien, Matthias Russ, Mark Fitzgerald and Wing Kong Chiu
Sensors 2022, 22(18), 6727; https://doi.org/10.3390/s22186727 - 6 Sep 2022
Viewed by 1555
Abstract
Osseointegration implant has attracted significant attention as an alternative treatment for transfemoral amputees. It has been shown to improve patients’ sitting and walking comfort and control of the artificial limb, compared to the conventional socket device. However, the patients treated with osseointegration implants [...] Read more.
Osseointegration implant has attracted significant attention as an alternative treatment for transfemoral amputees. It has been shown to improve patients’ sitting and walking comfort and control of the artificial limb, compared to the conventional socket device. However, the patients treated with osseointegration implants require a long rehabilitation period to establish sufficient femur–implant connection, allowing the full body weight on the prosthesis stem. Hence, a robust assessment method on the osseointegration process is essential to shorten the rehabilitation period and identify the degree of osseointegration prior to the connection of an artificial limb. This paper investigates the capability of a vibration-related index (E-index) on detecting the degree of simulated osseointegration process with three lengths of the residual femur (152, 190 and 228 mm). The adhesive epoxy with a setting time of 5 min was applied at the femur–implant interface to represent the stiffness change during the osseointegration process. The cross-spectrum and colormap of the normalised magnitude demonstrated significant changes during the cure time, showing that application of these plots could improve the accuracy of the currently available diagnostic techniques. Furthermore, the E-index exhibited a clear trend with a noticeable average increase of 53% against the cure time for all three residual length conditions. These findings highlight that the E-index can be employed as a quantitative justification to assess the degree of osseointegration process without selecting and tracing the resonant frequency based on the geometry of the residual femur. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Geometry of the novel osseointegration implant: (<b>a</b>) developed based on Patent US20200188140 and (<b>b</b>) modified oval-shape implant with hollow IM stem for the experiment, dimension in mm.</p> Full article ">Figure 2
<p>Cross-section of femur–implant interface with adhesive epoxy.</p> Full article ">Figure 3
<p>Two-sensor (S1 and S2) setup for composite femur model with markers for three residual length conditions.</p> Full article ">Figure 4
<p>Coherence function for residual length of 152 mm.</p> Full article ">Figure 5
<p>Coherence function for residual length of 190 mm.</p> Full article ">Figure 6
<p>Coherence function for residual length of 228 mm.</p> Full article ">Figure 7
<p>Cross-spectrum of normalised magnitude for the residual length of 152 mm at 0, 150, 300, 600, and 1140 s cure times.</p> Full article ">Figure 8
<p>Quality factor of selected resonant peaks as a function of time.</p> Full article ">Figure 9
<p>Cross-spectrum of normalised magnitude at 0, 150, 300, 600, and 1140 s cure times for residual lengths of (<b>a</b>) 190 and (<b>b</b>) 228 mm.</p> Full article ">Figure 10
<p>Quality factor of selected resonant peaks as function of time for residual lengths of (<b>a</b>) 190 and (<b>b</b>) 228 mm.</p> Full article ">Figure 11
<p>Colormap of the normalised magnitude development as a function of cure time for 152 mm.</p> Full article ">Figure 12
<p>Colormap of the normalised magnitude development as a function of cure time for 190 mm.</p> Full article ">Figure 13
<p>Colormap of the normalised magnitude development as a function of cure time for 228 mm.</p> Full article ">Figure 14
<p>E-index development as the function of cure time for 152 mm.</p> Full article ">Figure 15
<p>E-index development as the function of cure time for 190 mm.</p> Full article ">Figure 16
<p>E-index development as the function of cure time for 228 mm.</p> Full article ">Figure 17
<p>Average E-index development as the function of cure time for three residual lengths.</p> Full article ">
24 pages, 4128 KiB  
Article
An Adaptive and Robust Control Strategy for Real-Time Hybrid Simulation
by Hong-Wei Li, Fang Wang, Yi-Qing Ni, You-Wu Wang and Zhao-Dong Xu
Sensors 2022, 22(17), 6569; https://doi.org/10.3390/s22176569 - 31 Aug 2022
Cited by 7 | Viewed by 1679
Abstract
A real-time hybrid simulation (RTHS) is a promising technique to investigate a complicated or large-scale structure by dividing it into numerical and physical substructures and conducting cyber-physical tests on it. The control system design of an RTHS is a challenging topic due to [...] Read more.
A real-time hybrid simulation (RTHS) is a promising technique to investigate a complicated or large-scale structure by dividing it into numerical and physical substructures and conducting cyber-physical tests on it. The control system design of an RTHS is a challenging topic due to the additional feedback between the physical and numerical substructures, and the complexity of the physical control plant. This paper proposes a novel RTHS control strategy by combining the theories of adaptive control and robust control, where a reformed plant which is highly simplified compared to the physical plant can be used to design the control system without compromising the control performance. The adaptation and robustness features of the control system are realized by the bounded-gain forgetting least-squares estimator and the sliding mode controller, respectively. The control strategy is validated by investigating an RTHS benchmark problem of a nonlinear three-story steel frame The proposed control strategy could simplify the control system design and does not require a precise physical plant; thus, it is an efficient and practical option for an RTHS. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Tracking results for the forced Van der Pol oscillator.</p> Full article ">Figure 2
<p>True values and estimations of the original plant’s parameters.</p> Full article ">Figure 3
<p>Estimations of the reformed plant 1’s parameters.</p> Full article ">Figure 4
<p>Estimations of the reformed plant 2’s parameters.</p> Full article ">Figure 5
<p>RTHS block diagram.</p> Full article ">Figure 6
<p>RTHS partitioning scheme.</p> Full article ">Figure 7
<p>Force–displacement and force–velocity responses of the MR damper under sinusoidal displacement input: <math display="inline"><semantics> <mrow> <mn>5</mn> <mspace width="0.166667em"/> <mo form="prefix">sin</mo> <mrow> <mn>2.5</mn> <mi>π</mi> <mi>t</mi> </mrow> </mrow> </semantics></math> (mm).</p> Full article ">Figure 8
<p>Third-floor displacement responses of Intact w/o d and Intact w/ d (El Centro earthquake, Case 4).</p> Full article ">Figure 9
<p>Third-floor displacement responses of Intact w/ d and Damaged w/ d (El Centro earthquake, Case 4).</p> Full article ">Figure 10
<p>Block diagram of the physical plant for Intact w/ d.</p> Full article ">Figure 11
<p>Frequency responses of physical and reformed plants (<b>left</b>: nominal models; <b>right</b>: models with uncertainties).</p> Full article ">Figure 12
<p>Frequency response of the low-pass filter: <math display="inline"><semantics> <mrow> <msub> <mi>z</mi> <mn>1</mn> </msub> <mo>=</mo> <mn>1600</mn> <mspace width="0.166667em"/> <msup> <mi mathvariant="normal">s</mi> <mrow> <mo>−</mo> <mn>2</mn> </mrow> </msup> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <msub> <mi>z</mi> <mn>2</mn> </msub> <mo>=</mo> <mn>10</mn> <mspace width="0.166667em"/> <msup> <mi mathvariant="normal">s</mi> <mrow> <mo>−</mo> <mn>1</mn> </mrow> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 13
<p>Block diagram of the control system in RTHS.</p> Full article ">Figure 14
<p>Third-floor displacement responses of Intact w/o d (El Centro earthquake, Case 4, perturbed plant).</p> Full article ">Figure 15
<p>Third-floor displacement responses of Intact w/ d (El Centro earthquake, Case 4, perturbed plant).</p> Full article ">Figure 16
<p>Third-floor displacement responses of Damaged w/ d (El Centro earthquake, Case 4, perturbed plant).</p> Full article ">Figure 17
<p>Parameter estimations for Intact w/o d, Intact w/ d and Damaged w/d (El Centro earthquake, Case 4, perturbed plant).</p> Full article ">
22 pages, 12532 KiB  
Article
Convolutional Neural Network-Based Rapid Post-Earthquake Structural Damage Detection: Case Study
by Edisson Alberto Moscoso Alcantara and Taiki Saito
Sensors 2022, 22(17), 6426; https://doi.org/10.3390/s22176426 - 25 Aug 2022
Cited by 3 | Viewed by 1940
Abstract
It is necessary to detect the structural damage condition of essential buildings immediately after an earthquake to identify safe structures, evacuate, or resume crucial activities. For this reason, a CNN methodology proposed to detect the structural damage condition of a building is here [...] Read more.
It is necessary to detect the structural damage condition of essential buildings immediately after an earthquake to identify safe structures, evacuate, or resume crucial activities. For this reason, a CNN methodology proposed to detect the structural damage condition of a building is here improved and validated for two currently instrumented essential buildings (Tahara City Hall and Toyohashi Fire Station). Three-dimensional frames instead of lumped mass models are used for the buildings. Besides this, a methodology to select records is introduced to reduce the variability of the structural responses. The maximum inter-storey drift and absolute acceleration of each storey are used as damage indicators. The accuracy is evaluated by the usability of the building, total damage condition, storey damage condition, and total comparison of the damage indicators. Finally, the maximum accuracy and R2 of the responses are obtained as follows: for the Tahara City Hall building, 90.0% and 0.825, respectively; for the Toyohashi Fire Station building, 100% and 0.909, respectively. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Methodology flowchart to obtain the Convolutional Neural Network model.</p> Full article ">Figure 2
<p>Tahara City Hall Building.</p> Full article ">Figure 3
<p>General drawings of Tahara City Hall building. (<b>a</b>) Plan of 1st storey view. (<b>b</b>) Plan of 2nd and 3rd stories’ views. (<b>c</b>) Plan from 4th to 6th storey view. (<b>d</b>) Elevation of X-direction view.</p> Full article ">Figure 4
<p>Toyohashi Fire Station Building.</p> Full article ">Figure 5
<p>General drawings of Toyohashi Fire Station building. (<b>a</b>) Plan of basement view. (<b>b</b>) Plan from 1st to 6th storey view. (<b>c</b>) Plan of 7th storey view. (<b>d</b>) Elevation of X-direction view.</p> Full article ">Figure 5 Cont.
<p>General drawings of Toyohashi Fire Station building. (<b>a</b>) Plan of basement view. (<b>b</b>) Plan from 1st to 6th storey view. (<b>c</b>) Plan of 7th storey view. (<b>d</b>) Elevation of X-direction view.</p> Full article ">Figure 6
<p>Beam model with nonlinear flexural and shear springs [<a href="#B23-sensors-22-06426" class="html-bibr">23</a>].</p> Full article ">Figure 7
<p>Hysteresis model. (<b>a</b>) Degrading trilinear slip model for RC sections; (<b>b</b>) bilinear model for steel sections [<a href="#B23-sensors-22-06426" class="html-bibr">23</a>].</p> Full article ">Figure 8
<p>(<b>a</b>) Column model with multi-springs to consider Nz-Mx-My nonlinear interaction; (<b>b</b>) concrete and steel springs; (<b>c</b>) hysteresis model for steel and concrete springs [<a href="#B23-sensors-22-06426" class="html-bibr">23</a>].</p> Full article ">Figure 9
<p>Selection of ground motion records for inter-storey drift and acceleration flowchart.</p> Full article ">Figure 10
<p>IDA Curves of Tahara City Hall Building. (<b>a</b>) IDA curves of the database for SD; (<b>b</b>) IDA curves of selected records for SD; (<b>c</b>) IDA curves of the database for AA; (<b>d</b>) IDA curves of selected records for AA.</p> Full article ">Figure 11
<p>IDA Curves of Toyohashi Fire Station Building. (<b>a</b>) IDA curves of the database for SD; (<b>b</b>) IDA curves of selected records for SD; (<b>c</b>) IDA curves of the database for AA; (<b>d</b>) IDA curves of selected records for AA.</p> Full article ">Figure 11 Cont.
<p>IDA Curves of Toyohashi Fire Station Building. (<b>a</b>) IDA curves of the database for SD; (<b>b</b>) IDA curves of selected records for SD; (<b>c</b>) IDA curves of the database for AA; (<b>d</b>) IDA curves of selected records for AA.</p> Full article ">Figure 12
<p>Acceleration response spectrum of Tahara City Hall Building at T<sub>1</sub> = 0.681 s and Sa(T<sub>1</sub>) = 100 gal. (<b>a</b>) Training records for SD; (<b>b</b>) Training records for AA; (<b>c</b>) Validation records for SD and AA analyses.</p> Full article ">Figure 13
<p>Acceleration response spectrum of Toyohashi Fire Station Building at T<sub>1</sub> = 0.748 s and Sa(T<sub>1</sub>) = 100 gal. (<b>a</b>) Training records for SD; (<b>b</b>) training records for AA; (<b>c</b>) validation records for SD and AA analyses.</p> Full article ">Figure 14
<p>(<b>a</b>) Original signal; (<b>b</b>) two-dimensional WS; (<b>c</b>) two-dimensional WPS; (<b>d</b>) three-dimensional WPS; (<b>e</b>) three-dimensional WPS.</p> Full article ">Figure 14 Cont.
<p>(<b>a</b>) Original signal; (<b>b</b>) two-dimensional WS; (<b>c</b>) two-dimensional WPS; (<b>d</b>) three-dimensional WPS; (<b>e</b>) three-dimensional WPS.</p> Full article ">Figure 15
<p>A general CNN model.</p> Full article ">Figure 16
<p>Convolutional layer.</p> Full article ">Figure 17
<p>SD results of the TP and VP for Tahara City Hall building: (<b>a</b>) Convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of SD.</p> Full article ">Figure 18
<p>AA results of the TP and VP for Tahara City Hall building: (<b>a</b>) convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of AA.</p> Full article ">Figure 18 Cont.
<p>AA results of the TP and VP for Tahara City Hall building: (<b>a</b>) convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of AA.</p> Full article ">Figure 19
<p>SD results of the TP and VP for Toyohashi Fire Station building: (<b>a</b>) convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of SD.</p> Full article ">Figure 19 Cont.
<p>SD results of the TP and VP for Toyohashi Fire Station building: (<b>a</b>) convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of SD.</p> Full article ">Figure 20
<p>AA results of the TP and VP for Toyohashi Fire Station building: (<b>a</b>) convergence curve—loss in the TP; (<b>b</b>) confusion matrix—usability of the building by VP; (<b>c</b>) confusion matrix—total damage condition by VP; (<b>d</b>) confusion matrix—storey damage condition by VP; (<b>e</b>) total comparison of AA.</p> Full article ">
14 pages, 4494 KiB  
Article
Particle Swarm Optimization Algorithm for Guided Waves Based Damage Localization Using Fiber Bragg Grating Sensors in Remote Configuration
by Rohan Soman, Alex Boyer, Jee Myung Kim and Kara Peters
Sensors 2022, 22(16), 6000; https://doi.org/10.3390/s22166000 - 11 Aug 2022
Cited by 5 | Viewed by 1905
Abstract
Structural health monitoring (SHM) systems may allow a reduction in maintenance costs and extend the lifetime of the structure. As a result, they are of interest to the research community. Ideally, the SHM methods should be low cost, while being able to detect [...] Read more.
Structural health monitoring (SHM) systems may allow a reduction in maintenance costs and extend the lifetime of the structure. As a result, they are of interest to the research community. Ideally, the SHM methods should be low cost, while being able to detect and localize small levels of damage reliably and accurately. The fiber Bragg grating (FBG) sensors are light in weight, insensitive to electric and magnetic fields, and can be embedded. The edge filtering configuration for transduction allows the use of FBG for guided wave (GW) sensing. This sensitivity may be further enhanced through their application in the remote bonded configuration. This paper provides a proof-of-concept for the use of remotely bonded FBG for damage localization. In order to improve the computational efficiency, a particle swarm optimization (PSO) based algorithm is developed. The PSO allows a significant improvement in the computation time which makes it better suited for real-time damage localization. The proposed objective function is based on the exponential elliptical approach. First, the suitability of the PSO for damage localization is shown. Then the performance of the chosen objective function is compared with the brute-force algorithm as well as other objective functions found in the literature. The methodology is employed on a simple aluminum plate. The results indicate that indeed the objective function along with the PSO is suitable for damage localization. Also as the objective function is developed taking into consideration the specific challenges with the use of FBG sensors, performs better than the other objective functions as well as the brute force algorithm. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) Remote bonding configuration (<b>b</b>) Mode conversion in remotely bonded FBG.</p> Full article ">Figure 2
<p>Particle swarm Optimization Flow Chart based on [<a href="#B22-sensors-22-06000" class="html-bibr">22</a>].</p> Full article ">Figure 3
<p>Experiment Setup. 1—Tunable Laser, 2—Photodetector, 3—Power Supply, 4—Waveform Generator, 5—Voltage Amplifier, 6—Oscilloscope, 7—Experiment Sample. 8—Actuators (A1, A2, A3), 9—Sensors (S1, S2, S3).</p> Full article ">Figure 4
<p>Schematic of the experimental setup (Coordinates are in cm).</p> Full article ">Figure 5
<p>Convergence of the PSO to global maxima.</p> Full article ">Figure 6
<p>Swarm locations at each iteration (global best marked with a filled blue circle).</p> Full article ">Figure 7
<p>Damage localization for 6 damage scenarios. (black circle shows the real location of magnet).</p> Full article ">Figure 8
<p>Damage localization for damage 7 for 100 PSO runs (red dotted circle indicates the acceptable region for localization).</p> Full article ">Figure 9
<p>Error in damage localization for damage 7 for 100 PSO runs.</p> Full article ">Figure 10
<p>Damage maps with brute force method (<b>a</b>) D1 (<b>b</b>) D4.</p> Full article ">Figure 11
<p>Damage maps comparing the two cost functions (<b>a</b>) Number of ellipses for D2 (<b>b</b>) Exponential function for D2 (<b>c</b>) Number of ellipses for D3 (<b>d</b>) Exponential function for D3.</p> Full article ">
22 pages, 4315 KiB  
Article
Low-Cost Wireless Structural Health Monitoring of Bridges
by Seyedmilad Komarizadehasl, Fidel Lozano, Jose Antonio Lozano-Galant, Gonzalo Ramos and Jose Turmo
Sensors 2022, 22(15), 5725; https://doi.org/10.3390/s22155725 - 30 Jul 2022
Cited by 28 | Viewed by 5465
Abstract
Nowadays, low-cost accelerometers are getting more attention from civil engineers to make Structural Health Monitoring (SHM) applications affordable and applicable to a broader range of structures. The present accelerometers based on Arduino or Raspberry Pi technologies in the literature share some of the [...] Read more.
Nowadays, low-cost accelerometers are getting more attention from civil engineers to make Structural Health Monitoring (SHM) applications affordable and applicable to a broader range of structures. The present accelerometers based on Arduino or Raspberry Pi technologies in the literature share some of the following drawbacks: (1) high Noise Density (ND), (2) low sampling frequency, (3) not having the Internet’s timestamp with microsecond resolution, (4) not being used in experimental eigenfrequency analysis of a flexible and a less-flexible bridge, and (5) synchronization issues. To solve these problems, a new low-cost triaxial accelerometer based on Arduino technology is presented in this work (Low-cost Adaptable Reliable Accelerometer—LARA). Laboratory test results show that LARA has a ND of 51 µg/√Hz, and a frequency sampling speed of 333 Hz. In addition, LARA has been applied to the eigenfrequency analysis of a short-span footbridge and its results are compared with those of a high-precision commercial sensor. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>LARA elements: (<b>a</b>) the adjustments and wire connections of the sensing part, (<b>b</b>) the sensing and acquisition part, and (<b>c</b>) LARA in detail.</p> Full article ">Figure 2
<p>Frequency domain diagrams for <span class="html-italic">z</span> axis of: (<b>a</b>) MPU9250, (<b>b</b>) CHEAP, and (<b>c</b>) LARA.</p> Full article ">Figure 3
<p>Frequency domain diagrams of LARA for: (<b>a</b>) <span class="html-italic">z</span>, (<b>b</b>) <span class="html-italic">x,</span> and (<b>c</b>) <span class="html-italic">y</span>-axis.</p> Full article ">Figure 4
<p>Laboratory validation of LARA: (<b>a</b>) mounting CHEAP and LARA to the shaking part of the jack, and (<b>b</b>) the used vibrating platform (INSTRON 8803).</p> Full article ">Figure 5
<p>FFT representation of the low-frequency signals: (<b>a</b>) 0.1 Hz, (<b>b</b>) 0.2 Hz, (<b>c</b>) 0.3 Hz, and (<b>d</b>) 0.5 Hz.</p> Full article ">Figure 6
<p>Displacement report of the jack in a frequency-domain diagram.</p> Full article ">Figure 7
<p>Displacement report of the accelerometers: (<b>a</b>) MPU9250, (<b>b</b>) CHEAP, and (<b>c</b>) LARA.</p> Full article ">Figure 8
<p>The time-domain presentation of a vibration acquisition with RMS value of one g by LARA.</p> Full article ">Figure 9
<p>The time-domain presentation of acceleration amplitude saturation of LARA.</p> Full article ">Figure 10
<p>(<b>a</b>) A picture of the pass way, (<b>b</b>) plan of the bridge, and (<b>c</b>) section of the pass way bridge (all units are in mm).</p> Full article ">Figure 11
<p>Mounting the sensors to the mid span of the bridge under study: (<b>a</b>) mounting diagram of LARA to the bridge and (<b>b</b>) photo of LARA and HI-INC sensor mounted on a footbridge in Barcelona.</p> Full article ">Figure 12
<p>Eigenfrequency analysis of a footbridge using LARA for (<b>a</b>) vertical, (<b>b</b>) longitudinal, and (<b>c</b>) transversal directions of the footbridge.</p> Full article ">
18 pages, 4222 KiB  
Article
Definition of Optimized Indicators from Sensors Data for Damage Detection of Instrumented Roadways
by David Souriou, Jean-Michel Simonin and Franziska Schmidt
Sensors 2022, 22(15), 5572; https://doi.org/10.3390/s22155572 - 26 Jul 2022
Viewed by 1604
Abstract
Pavement instrumentation sensors have become a common practice for structural monitoring. Data processing often consists of extracting different values generally representative of variations in the whole pavement, but this makes it more difficult to follow a specific characteristic of the structure. To overcome [...] Read more.
Pavement instrumentation sensors have become a common practice for structural monitoring. Data processing often consists of extracting different values generally representative of variations in the whole pavement, but this makes it more difficult to follow a specific characteristic of the structure. To overcome this limitation, this paper aimed to transpose an original method, labeled the Optimized Indicators Method (OIM), which consists of finding some weighting functions of a signal to evaluate some indicators linked to a physical characteristic of the structure. The main advantage of this method is that the inferred indicators are particularly sensitive to a specific characteristic of the structure without being sensitive to the others. This research work consisted of analyzing, conventionally and with the OIM, the signals of strain gauges, which were recorded during an experimental campaign carried out on a circular instrumented pavement and submitted to an accelerated fatigue testing. The OIM, calibrated through an experimental reference signal, makes it possible to evaluate independently, through weighting functions, some specific physical characteristics of the pavement. It showed a two-stepped degradation, starting with a damaging of the bituminous layer, followed by an alteration of the base layer, which could not be easily deduced from a conventional processing method. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Flowchart for the determination of weighting functions and optimized indicators of a structure parameter.</p> Full article ">Figure 2
<p>The pavement fatigue carousel at Univ. Eiffel Nantes [<a href="#B16-sensors-22-05572" class="html-bibr">16</a>].</p> Full article ">Figure 3
<p>(<b>a</b>) Schematic representation in front view of the pavement’s structure, (<b>b</b>) top view illustrating the orientation of the strain gauge along the dual-wheel load direction, (<b>c</b>) picture of a DYNATEST<sup>®</sup> strain gauge.</p> Full article ">Figure 4
<p>Example of µ strain signals recorded from the sensor as a function of time during spinning charging cycles.</p> Full article ">Figure 5
<p>Mean temperatures and date of acquisition recorded inside the bituminous layer as a function of the number of loads.</p> Full article ">Figure 6
<p>Huet–Sayegh model [<a href="#B22-sensors-22-05572" class="html-bibr">22</a>].</p> Full article ">Figure 7
<p>Fitting of the proposed numerical model to the reference signal.</p> Full article ">Figure 8
<p>Fitting of the proposed numerical model to the reference signal with (<b>a</b>) ±10% variation for <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mo>∞</mo> </msub> </mrow> </semantics></math>, (<b>b</b>) ±10% variation for <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mrow> <mi>U</mi> <mi>G</mi> <mi>M</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 9
<p>Weighting functions as a function of distance of <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mo>∞</mo> </msub> </mrow> </semantics></math> (blue axis in the center) and <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mrow> <mi>U</mi> <mi>G</mi> <mi>M</mi> </mrow> </msub> </mrow> </semantics></math> (red axis in the right) obtained from the reference signal.</p> Full article ">Figure 10
<p>Evolution of gauge signals with the loading, from 151,200 loadings (beginning of the test) to 881,600 loadings (end of test).</p> Full article ">Figure 11
<p>Relative evolution of the proposed indicators with the number of loads.</p> Full article ">Figure 12
<p>Relative evolution of the optimized indicators with the number of loads.</p> Full article ">Figure 13
<p>Evaluation of <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mo>∞</mo> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mrow> <mi>U</mi> <mi>G</mi> <mi>M</mi> </mrow> </msub> </mrow> </semantics></math> as a function of loadings through the OIM with recalibration steps proceeding at numbers of loadings labeled as N<sub>ref</sub> with ±10% variation windows (<b>a</b>) for <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mo>∞</mo> </msub> </mrow> </semantics></math> and (<b>b</b>) for <math display="inline"><semantics> <mrow> <msub> <mi>E</mi> <mrow> <mi>U</mi> <mi>G</mi> <mi>M</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">
23 pages, 12605 KiB  
Article
A Study on the Applicability of the Impact-Echo Test Using Semi-Supervised Learning Based on Dynamic Preconditions
by Young-Geun Yoon, Chung-Min Kim and Tae-Keun Oh
Sensors 2022, 22(15), 5484; https://doi.org/10.3390/s22155484 - 22 Jul 2022
Cited by 10 | Viewed by 2703
Abstract
The Impact-Echo (IE) test is an effective method for determining the presence, depth, and area of cracks in concrete as well as the dimensions of the sound concrete without defects. In addition, shallow delamination can be measured by confirming a flexural mode in [...] Read more.
The Impact-Echo (IE) test is an effective method for determining the presence, depth, and area of cracks in concrete as well as the dimensions of the sound concrete without defects. In addition, shallow delamination can be measured by confirming a flexural mode in the low-frequency region. Owing to the advancement of non-contact sensors and automated measurement equipment, the IE test can be measured at multiple points in a short period. To analyze and distinguish a large volume of data, applying supervised learning (SL) associated with various contemporary algorithms is necessary. However, SL has limitations due to the difficulty in accurate labeling for increased volumes of test data, and reflection of new specimen characteristics, and it is necessary to apply semi-supervised learning (SSL) to overcome them. This study analyzes the accuracy and evaluates the applicability of a model trained with SSL rather than SL using the data from the air-coupled IE test based on dynamic preconditions. For the detection of delamination defects, the dynamic behavior-based flexural mode was identified, and 21 features were extracted in the time and frequency domains. Three principal components (PCs) such as the real moment, real RMS, and imaginary moment were derived through principal component analysis (PCA). PCs were identical in slab, pavement, and deck. In the case of SSL considering a dynamic behavior, the accuracy increased by 7–8% compared with SL, and it could categorize good, fair, and poor status to a higher level for actual structures. The applicability of SSL to the IE test was confirmed, and because the crack progress varies under field conditions, other parameters must be considered in the future to reflect this. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic of Impact-Echo method.</p> Full article ">Figure 2
<p>Criteria governing thick plate, thin plate, and membrane cases based on the side-to-thickness ratio <span class="html-italic">a</span>/<span class="html-italic">h</span> (<span class="html-italic">a</span> &gt; <span class="html-italic">b</span>). (<b>a</b>) The plate theory range according to <span class="html-italic">a</span>/<span class="html-italic">h</span>; (<b>b</b>) the tendency of thickness and flexural mode according to <span class="html-italic">a</span>/<span class="html-italic">h</span>.</p> Full article ">Figure 3
<p>PCA procedure.</p> Full article ">Figure 4
<p>General framework of the self-training classifier.</p> Full article ">Figure 5
<p>Test specimens for training and validation with SL and SSL.</p> Full article ">Figure 5 Cont.
<p>Test specimens for training and validation with SL and SSL.</p> Full article ">Figure 6
<p>Test equipment and procedure: (<b>a</b>) test equipment; (<b>b</b>) test procedure for slab A; (<b>c</b>) test procedure for slab B; (<b>d</b>) test procedure for slab C.</p> Full article ">Figure 7
<p>Representative spectrum by defect type and slab.</p> Full article ">Figure 8
<p>Signal processing and feature extraction procedure for PCA.</p> Full article ">Figure 9
<p>Three-dimensional plot of three principal components using PCA for each slab: (<b>a</b>) slab A, (<b>b</b>) slab B, (<b>c</b>) slab C.</p> Full article ">Figure 10
<p>Predicted region and confusion chart through the SL model: (<b>a</b>) 2D plots of the full data used for training and validation; (<b>b</b>) regions of three labels predicted by SL; (<b>c</b>) prediction accuracy of the SL model.</p> Full article ">Figure 10 Cont.
<p>Predicted region and confusion chart through the SL model: (<b>a</b>) 2D plots of the full data used for training and validation; (<b>b</b>) regions of three labels predicted by SL; (<b>c</b>) prediction accuracy of the SL model.</p> Full article ">Figure 11
<p>SSL model accuracy analysis by combination.</p> Full article ">Figure 12
<p>First performance verification of the SL and SSL models: (<b>a</b>) DL-A2, (<b>b</b>) DL-A4, (<b>c</b>) DL-A5.</p> Full article ">Figure 13
<p>C-scan images for IE data collected from the bridge deck; spectral data are displayed up to 6 kHz. The closed and open circles are the positions of the good and poor cores, respectively. (<b>a</b>) Overlapped images of four NDT results analyzed by experts (image courtesy: Taekeun Oh), (<b>b</b>) prediction result of the SL model, (<b>c</b>) prediction result of the SSL model.</p> Full article ">Figure 14
<p>Eight drilled core samples, C3, C4, and C8, contain horizontal delamination at the top bar depth (image credit: Nenad Gucunski).</p> Full article ">Figure 15
<p>Proposed flowchart to optimize the IE data of concrete specimens.</p> Full article ">
30 pages, 9618 KiB  
Article
Deep Learning Empowered Structural Health Monitoring and Damage Diagnostics for Structures with Weldment via Decoding Ultrasonic Guided Wave
by Zi Zhang, Hong Pan, Xingyu Wang and Zhibin Lin
Sensors 2022, 22(14), 5390; https://doi.org/10.3390/s22145390 - 19 Jul 2022
Cited by 15 | Viewed by 2845
Abstract
Welding is widely used in the connection of metallic structures, including welded joints in oil/gas metallic pipelines and other structures. The welding process is vulnerable to the inclusion of different types of welding defects, such as lack of penetration and undercut. These defects [...] Read more.
Welding is widely used in the connection of metallic structures, including welded joints in oil/gas metallic pipelines and other structures. The welding process is vulnerable to the inclusion of different types of welding defects, such as lack of penetration and undercut. These defects often initialize early-age cracking and induced corrosion. Moreover, welding-induced defects often accompany other types of mechanical damage, thereby leading to more challenges in damage detection. As such, identification of weldment defects and interaction with other mechanical damages at their early stage is crucial to ensure structural integrity and avoid potential premature failure. The current strategies of damage identification are achieved using ultrasonic guided wave approaches that rely on a change in physical parameters of propagating waves to discriminate as to whether there exist damaged states or not. However, the inherently complex nature of weldment, the complication of damages interactions, and large-scale/long span structural components integrated with structure uncertainties pose great challenges in data interpretation and making an informed decision. Artificial intelligence and machine learning have recently become emerging methods for data fusion, with great potential for structural signal processing through decoding ultrasonic guided waves. Therefore, this study aimed to employ the deep learning method, convolutional neural network (CNN), for better characterization of damage features in terms of welding defect type, severity, locations, and interaction with other damage types. The architecture of the CNN was set up to provide an effective classifier for data representation and data fusion. A total of 16 damage states were designed for training and calibrating the accuracy of the proposed method. The results revealed that the deep learning method enables effectively and automatically extracting features of ultrasonic guided waves and yielding high precise prediction for damage detection of structures with welding defects in complex situations. In addition, the effectiveness and robustness of the proposed methods for structure uncertainties using different embedding materials, and data under noise interference, was also validated and findings demonstrated that the proposed deep learning methods still exhibited a high accuracy at high noise levels. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Flow of the proposed concept in this study.</p> Full article ">Figure 2
<p>Phase velocity and group velocity. (<b>a</b>) Phase velocity. (<b>b</b>) Group velocity.</p> Full article ">Figure 3
<p>Excited guided waves used in the simulation.</p> Full article ">Figure 4
<p>Pipe model from the literature.</p> Full article ">Figure 5
<p>Comparison of the results to those from the literature [<a href="#B44-sensors-22-05390" class="html-bibr">44</a>].</p> Full article ">Figure 6
<p>COMSOL models of pipes in different states: (<b>a</b>) intact pipe; (<b>b</b>) pipe with a weld; (<b>c</b>) pipe with a crack; (<b>d</b>) pipe with crack and weld.</p> Full article ">Figure 7
<p>Signal characteristics of the pipes under varying cases: (<b>a</b>) intact pipe; (<b>b</b>) pipe with a weld; (<b>c</b>) pipe with a crack; (<b>d</b>) pipe with crack and weld.</p> Full article ">Figure 8
<p>Flowchart for damage detection using the CNN.</p> Full article ">Figure 9
<p>Flowchart for the damage detection by CNN.</p> Full article ">Figure 10
<p>Convolutional layer.</p> Full article ">Figure 11
<p>Pooling layer.</p> Full article ">Figure 12
<p>FE modeling of a pipe. (<b>a</b>) Steel pipe with a welded joint and notch-shaped damage. (<b>b</b>) Excitation nodes. (<b>c</b>) V-shaped weldment.</p> Full article ">Figure 13
<p>Welding defects (<b>a</b>) Lack of fusion; (<b>b</b>) Cracks; (<b>c</b>) Undercut; (<b>d</b>) Lack of penetration.</p> Full article ">Figure 14
<p>Wave propagation through the entire span of the pipe: (<b>a</b>–<b>f</b>).</p> Full article ">Figure 15
<p>Signals with different noise levels. (<b>a</b>) Original signal; (<b>b</b>) SNR = 100 dB; (<b>c</b>) SNR = 80 dB; (<b>d</b>) SNR = 60 dB.</p> Full article ">Figure 16
<p>Weights of the third convolutional layer at 60 dB: (<b>a</b>) Group 1; (<b>b</b>) Group 2; (<b>c</b>) Group 3; (<b>d</b>) Group 4; (<b>e</b>) Group 5.</p> Full article ">Figure 17
<p>Feature maps. (<b>a</b>) SNR = 100 dB; (<b>b</b>) SNR = 60 dB.</p> Full article ">Figure 18
<p>Training and validation results under noise of. (<b>a</b>) SNR = 80 dB; (<b>b</b>) SNR = 70 dB; (<b>c</b>) SNR = 60 dB.</p> Full article ">Figure 19
<p>Training and validation results. (<b>a</b>) SNR = 80 dB; (<b>b</b>) SNR = 70 dB; (<b>c</b>) SNR = 60 dB.</p> Full article ">Figure 20
<p>Testing results. (<b>a</b>) SNR = 80 dB (Accuracy = 100%); (<b>b</b>) SNR = 70 dB (Accuracy = 92.3%).</p> Full article ">Figure 21
<p>Testing results. (<b>a</b>) SNR = 70 dB (Accuracy = 95%). (<b>b</b>) SNR = 60 dB (Accuracy = 68%).</p> Full article ">Figure 22
<p>Locations identification in pipeline. (<b>a</b>) Location detects without welding defect. (<b>b</b>) Location detects with welding defect.</p> Full article ">Figure 23
<p>Training and validation results. (<b>a</b>) SNR = 70 dB. (<b>b</b>) SNR = 60 dB.</p> Full article ">Figure 24
<p>Testing results. (<b>a</b>) SNR = 70 dB (Accuracy = 99.4%). (<b>b</b>) SNR = 60 dB (Accuracy = 75.4%).</p> Full article ">Figure 25
<p>Training and validation results. (<b>a</b>) SNR = 80 dB (<b>b</b>) SNR = 70 dB.</p> Full article ">Figure 26
<p>Confusion matrix associated with defect severity. (<b>a</b>) SNR = 80 dB (Accuracy = 100%). (<b>b</b>) SNR = 70 dB (Accuracy = 88.8%).</p> Full article ">Figure 27
<p>Accuracy with respect to defect severity under various noise levels.</p> Full article ">Figure 28
<p>Comparison of the proposed damage detection with other two methods in accuracy.</p> Full article ">Figure 29
<p>Model of a buried pipe.</p> Full article ">Figure 30
<p>Detectability for pipes under different embedment conditions: (<b>a</b>–<b>f</b>). (<b>a</b>) Pipe without embedment (Accuracy = 100%, SNR = 100 dB). (<b>b</b>) Pipe with embedding soil (Accuracy = 100%, SNR=100 dB). (<b>c</b>) Pipe with embedding concrete (Accuracy = 97.3%, SNR = 100 dB). (<b>d</b>) Pipe without embedment (Accuracy = 88.8%, SNR = 70 dB); (<b>e</b>) Pipe with embedding soil (Accuracy = 82%, SNR = 70 dB); (<b>f</b>) Pipe with embedding concrete (Accuracy = 42%, SNR = 70 dB).</p> Full article ">Figure 30 Cont.
<p>Detectability for pipes under different embedment conditions: (<b>a</b>–<b>f</b>). (<b>a</b>) Pipe without embedment (Accuracy = 100%, SNR = 100 dB). (<b>b</b>) Pipe with embedding soil (Accuracy = 100%, SNR=100 dB). (<b>c</b>) Pipe with embedding concrete (Accuracy = 97.3%, SNR = 100 dB). (<b>d</b>) Pipe without embedment (Accuracy = 88.8%, SNR = 70 dB); (<b>e</b>) Pipe with embedding soil (Accuracy = 82%, SNR = 70 dB); (<b>f</b>) Pipe with embedding concrete (Accuracy = 42%, SNR = 70 dB).</p> Full article ">Figure 31
<p>Accuracy of the models for embedding cases with respect to different noise levels.</p> Full article ">
22 pages, 12159 KiB  
Article
Detection of Partially Structural Collapse Using Long-Term Small Displacement Data from Satellite Images
by Alireza Entezami, Carlo De Michele, Ali Nadir Arslan and Bahareh Behkamal
Sensors 2022, 22(13), 4964; https://doi.org/10.3390/s22134964 - 30 Jun 2022
Cited by 10 | Viewed by 1889
Abstract
The development of satellite sensors and interferometry synthetic aperture radar (InSAR) technology has enabled the exploitation of their benefits for long-term structural health monitoring (SHM). However, some restrictions cause this process to provide a small number of images leading to the problem of [...] Read more.
The development of satellite sensors and interferometry synthetic aperture radar (InSAR) technology has enabled the exploitation of their benefits for long-term structural health monitoring (SHM). However, some restrictions cause this process to provide a small number of images leading to the problem of small data for SAR-based SHM. Conversely, the major challenge of the long-term monitoring of civil structures pertains to variations in their inherent properties by environmental and/or operational variability. This article aims to propose new hybrid unsupervised learning methods for addressing these challenges. The methods in this work contain three main parts: (i) data augmentation by the Markov Chain Monte Carlo algorithm, (ii) feature normalization, and (iii) decision making via Mahalanobis-squared distance. The first method presented in this work develops an artificial neural network-based feature normalization by proposing an iterative hyperparameter selection of hidden neurons of the network. The second method is a novel unsupervised teacher–student learning by combining an undercomplete deep neural network and an overcomplete single-layer neural network. A small set of long-term displacement samples extracted from a few SAR images of TerraSAR-X is applied to validate the proposed methods. The results show that the methods can effectively deal with the major challenges in the SAR-based SHM applications. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Flowchart of the proposed multi-stage unsupervised learning method.</p> Full article ">Figure 2
<p>The graphical representation of the unsupervised feature normalization by ANN.</p> Full article ">Figure 3
<p>(<b>a</b>) Deep feedforward neural network with undercomplete configuration; (<b>b</b>) single-layer neural network with overcomplete configuration.</p> Full article ">Figure 4
<p>The flowchart of the proposed TSL.</p> Full article ">Figure 5
<p>Sample neurons for hidden layers of the DNN.</p> Full article ">Figure 6
<p>(<b>a</b>) The Tadcaster Bridge and its collapsed area [<a href="#B17-sensors-22-04964" class="html-bibr">17</a>]; (<b>b</b>) the bridge after reconstruction of the damaged area; (<b>c</b>) the plan view, pier labels, and collapse area.</p> Full article ">Figure 7
<p>The eight points of displacement samples as designated by “a”, “b”, “c”, “d”, “e”, “f”, “g”, and “h” (i.e., point “b” is the location of the partial collapse) [<a href="#B17-sensors-22-04964" class="html-bibr">17</a>].</p> Full article ">Figure 8
<p>Displacement samples of the eight points [<a href="#B17-sensors-22-04964" class="html-bibr">17</a>].</p> Full article ">Figure 9
<p>Augmented displacement samples based on the MCMC and HMC sampler by using the 100 sample points.</p> Full article ">Figure 10
<p>Box plot of the augmented displacement samples.</p> Full article ">Figure 11
<p>Damage detection of the Tadcaster Bridge without feature normalization.</p> Full article ">Figure 12
<p>Hyperparameter selection of the auto-associative neural network.</p> Full article ">Figure 13
<p>Evaluation of the effect of the ANN-based feature normalization: (<b>a</b>) MCMC-MSD; (<b>b</b>) MCMC-ANN-MSD.</p> Full article ">Figure 14
<p>Damage detection of the Tadcaster Bridge using MCMC-ANN-MSD and the CLT-based threshold estimator.</p> Full article ">Figure 15
<p>Damage detection of the Tadcaster Bridge using the GOF-based threshold estimator.</p> Full article ">Figure 16
<p>Selection of the optimal number of neurons of the seven hidden layers of the DNN.</p> Full article ">Figure 17
<p>Evaluation of the effect of the TSL-based feature normalization: (<b>a</b>) MCMC-MSD; (<b>b</b>) MCMC-TSL-MSD.</p> Full article ">Figure 18
<p>Damage detection of the Tadcaster Bridge using MCMC-TSL-MSD and the CLT-based threshold estimator.</p> Full article ">
20 pages, 39395 KiB  
Article
Model-Assisted Guided-Wave-Based Approach for Disbond Detection and Size Estimation in Honeycomb Sandwich Composites
by Piotr Fiborek and Paweł Kudela
Sensors 2021, 21(24), 8183; https://doi.org/10.3390/s21248183 - 8 Dec 2021
Cited by 9 | Viewed by 2450
Abstract
One of the axioms of structural health monitoring states that the severity of damage assessment can only be done in a learning mode under the supervision of an expert. Therefore, a numerical analysis was conducted to gain knowledge regarding the influence of the [...] Read more.
One of the axioms of structural health monitoring states that the severity of damage assessment can only be done in a learning mode under the supervision of an expert. Therefore, a numerical analysis was conducted to gain knowledge regarding the influence of the damage size on the propagation of elastic waves in a honeycomb sandwich composite panel. Core-skin debonding was considered as damage. For this purpose, a panel was modelled taking into account the real geometry of the honeycomb core using the time-domain spectral element method and two-dimensional elements. The presented model was compared with the homogenized model of the honeycomb core and validated in the experimental investigation. The result of the parametric study is a function of the influence of damage on the amplitude and energy of propagating waves. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Non-matching interface setup: (<b>a</b>) interface coupling and (<b>b</b>) degrees-of-freedom of the interface and the substructures.</p> Full article ">Figure 2
<p>Experimental setup for the (1) SDLV measurement—dashed line and (2) PZT wave acquisition—solid line.</p> Full article ">Figure 3
<p>Sample configuration: (<b>a</b>) top view of the sample, (<b>b</b>) honeycomb sandwich substructures and (<b>c</b>) details of the honeycomb cell.</p> Full article ">Figure 4
<p>The mesh with the node distribution, (<b>a</b>) spectral element used for modeling the wall of the core, (<b>b</b>) excerpt of the skin plate and (<b>c</b>) cyanoacrylate glue mesh generated in GMSH.</p> Full article ">Figure 5
<p>The damaged area in the: (<b>a</b>) experimental sample and (<b>b</b>) numerical mesh.</p> Full article ">Figure 6
<p>A flowchart representing the process for damage size estimation.</p> Full article ">Figure 7
<p>(<b>a</b>) The sensor signal <math display="inline"><semantics> <mrow> <mi mathvariant="sans-serif">Ψ</mi> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </semantics></math> windowed by a flattened Gaussian window <math display="inline"><semantics> <mrow> <mi>g</mi> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </semantics></math> and (<b>b</b>) the damage size estimation from the MADIF.</p> Full article ">Figure 8
<p>The top surface out of plane particle velocity snapshots in time 100 <math display="inline"><semantics> <mi mathvariant="sans-serif">μ</mi> </semantics></math>s for (<b>a</b>) the experimental results obtained by using SLDV, (<b>b</b>) the present model and (<b>c</b>) the homogenized model in the pristine sample.</p> Full article ">Figure 9
<p>The top surface out of plane particle velocity snapshots in time 100 <math display="inline"><semantics> <mi mathvariant="sans-serif">μ</mi> </semantics></math>s for (<b>a</b>) the experimental results obtained by using SLDV, (<b>b</b>) the present model and (<b>c</b>) the homogenized model in the sample with 90 mm damage.</p> Full article ">Figure 10
<p>Relative change of the energy of the half of the first package in the function of damage size.</p> Full article ">Figure 11
<p>Relative change of the maximum amplitude of the first package in the function of damage size.</p> Full article ">Figure 12
<p>The mean absolute error of the indices.</p> Full article ">Figure 13
<p>The model-assisted damage identification function (MADIF).</p> Full article ">

Review

Jump to: Research

29 pages, 6132 KiB  
Review
Smartphone Prospects in Bridge Structural Health Monitoring, a Literature Review
by Ekin Ozer and Rolands Kromanis
Sensors 2024, 24(11), 3287; https://doi.org/10.3390/s24113287 - 21 May 2024
Cited by 2 | Viewed by 1833
Abstract
Bridges are critical components of transportation networks, and their conditions have effects on societal well-being, the economy, and the environment. Automation needs in inspections and maintenance have made structural health monitoring (SHM) systems a key research pillar to assess bridge safety/health. The last [...] Read more.
Bridges are critical components of transportation networks, and their conditions have effects on societal well-being, the economy, and the environment. Automation needs in inspections and maintenance have made structural health monitoring (SHM) systems a key research pillar to assess bridge safety/health. The last decade brought a boom in innovative bridge SHM applications with the rise in next-generation smart and mobile technologies. A key advancement within this direction is smartphones with their sensory usage as SHM devices. This focused review reports recent advances in bridge SHM backed by smartphone sensor technologies and provides case studies on bridge SHM applications. The review includes model-based and data-driven SHM prospects utilizing smartphones as the sensing and acquisition portal and conveys three distinct messages in terms of the technological domain and level of mobility: (i) vibration-based dynamic identification and damage-detection approaches; (ii) deformation and condition monitoring empowered by computer vision-based measurement capabilities; (iii) drive-by or pedestrianized bridge monitoring approaches, and miscellaneous SHM applications with unconventional/emerging technological features and new research domains. The review is intended to bring together bridge engineering, SHM, and sensor technology audiences with decade-long multidisciplinary experience observed within the smartphone-based SHM theme and presents exemplary cases referring to a variety of levels of mobility. Full article
(This article belongs to the Special Issue Advanced Sensing Technologies in Structural Health Monitoring and Its Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>A timeline of smartphone evolution from its advent to complex and large-scale SHM research (2007 to 2024), examples above refer to in [<a href="#B27-sensors-24-03287" class="html-bibr">27</a>,<a href="#B28-sensors-24-03287" class="html-bibr">28</a>,<a href="#B29-sensors-24-03287" class="html-bibr">29</a>,<a href="#B31-sensors-24-03287" class="html-bibr">31</a>,<a href="#B32-sensors-24-03287" class="html-bibr">32</a>,<a href="#B33-sensors-24-03287" class="html-bibr">33</a>,<a href="#B34-sensors-24-03287" class="html-bibr">34</a>,<a href="#B35-sensors-24-03287" class="html-bibr">35</a>,<a href="#B36-sensors-24-03287" class="html-bibr">36</a>,<a href="#B37-sensors-24-03287" class="html-bibr">37</a>,<a href="#B38-sensors-24-03287" class="html-bibr">38</a>,<a href="#B39-sensors-24-03287" class="html-bibr">39</a>,<a href="#B40-sensors-24-03287" class="html-bibr">40</a>,<a href="#B42-sensors-24-03287" class="html-bibr">42</a>], respectively.</p> Full article ">Figure 2
<p>Level of mobilities (LoMs) observed on a smartphone-engaged bridge monitoring ecosystem.</p> Full article ">Figure 3
<p>Spatial footprint comparison for LoMs.</p> Full article ">Figure 4
<p>Conceptual breakdown of smartphone components.</p> Full article ">Figure 5
<p>Bridge cable installation of smartphone sensor: smartphone vibration frequency comparisons with the reference data (<b>a</b>,<b>b</b>) and sensor configuration (<b>c</b>), Yu et al. (2015) [<a href="#B32-sensors-24-03287" class="html-bibr">32</a>].</p> Full article ">Figure 6
<p>Image frame with an ROI (<b>left</b>) and ROI with a target (<b>right</b>) [<a href="#B82-sensors-24-03287" class="html-bibr">82</a>].</p> Full article ">Figure 7
<p>A vehicle–bridge interaction model used for simulating the indirect response, Zhu and Malekjafarian (2019) [<a href="#B120-sensors-24-03287" class="html-bibr">120</a>].</p> Full article ">Figure 8
<p>CV-based system for measurement collection of the Wilford Suspension bridge.</p> Full article ">Figure 9
<p>A sketch of the Wilford Suspension bridge with the locations of targets (T<math display="inline"><semantics> <mrow> <mi>i</mi> </mrow> </semantics></math>, where <math display="inline"><semantics> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mo> </mo> <mn>2</mn> <mo>,</mo> <mo> </mo> <mo>…</mo> <mo>,</mo> <mo> </mo> <mn>9</mn> </mrow> </semantics></math>).</p> Full article ">Figure 10
<p>Vertical displacement time-history for the selected activity (<b>left</b>) and power spectrum density plot (<b>right</b>).</p> Full article ">Figure 11
<p>The mode shape of the bridge at its first vertical frequency.</p> Full article ">Figure 12
<p>(<b>a</b>) Mudd–Schapiro Bridge, (<b>b</b>) accelerometer instrumentation, (<b>c</b>) sample smartphone measurement, (<b>d</b>) its frequency spectrum, and (<b>e</b>) and the comparison of crowdsourcing-identified modal frequency values with reference tests [<a href="#B34-sensors-24-03287" class="html-bibr">34</a>].</p> Full article ">Figure 13
<p>Smartphone-based SHM frameworks for (<b>a</b>) spatiotemporally uncontrolled [<a href="#B55-sensors-24-03287" class="html-bibr">55</a>] and (<b>b</b>) directionally distorted [<a href="#B56-sensors-24-03287" class="html-bibr">56</a>] and (<b>c</b>) and indirectly retrieved vibration data [<a href="#B57-sensors-24-03287" class="html-bibr">57</a>].</p> Full article ">Figure 14
<p>(<b>a</b>) Overall distribution of smartphone-based bridge monitoring sample studies according to their LoM context; (<b>b</b>–<b>d</b>) yearly publication trends among LoM1, LoM2, and LoM3, respectively.</p> Full article ">Figure 15
<p>A vision of collective and connected smartphone applications for bridge SHM.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-28

Planned Papers

The below list represents only planned manuscripts. Some of these manuscripts have not been received by the Editorial Office yet. Papers submitted to MDPI journals are subject to peer-review.

Title: A UAV-based Structural Health Monitoring and Computer Vision-aided Procedures for Seismic Safety Measures of Linear Infrastructures
Authors: Luna Ngeljaratan 1,2, Elif Ecem Bas 1,3, and Mohamed A. Moustafa 1*
Affiliation: 1 Department of Civil and Environmental Engineering, University of Nevada, Reno, NV 89557-0258, United States 2 Research Center for Structural Strength Technology, National Research and Innovation Agency, Kawasan Puspiptek Gedung 220, Setu, Tangerang Selatan 15314, Indonesia 3 R&D Test Systems, Aarhus, Middle Jutland, Denmark
Abstract: Computer vision in the Structural Health Monitoring (SHM) field has become popular, especially for processing Unmanned Aerial Vehicle (UAV)-based SHM data but still has limitations either in experimental testing or in practical application. Prior works have been focusing on UAV challenges and opportunities for vibration-based SHM of buildings or bridges but the empirical gap exists specifically for linear infrastructures. Since they are critical for the transportation of products and transmission of energy, the feasibility study of UAV-based SHM for linear infrastructures is essential to ensure their service continuity through an advanced SHM system. Thus, this study proposes a single Unmanned Aerial Vehicle (UAV) for seismic monitoring and safety assessment of linear infrastructures along with their computer vision-aided procedures. Among linear infrastructures, condition or seismic monitoring of natural gas pipelines is slightly more critical since their network may be constructed through several terrains with different seismicity leading to different hazard exposure. Therefore, the proposed procedures are implemented on natural gas pipelines under a large-scale shake-table test. The objective is to implement robust feature detection, extraction, and matching algorithms on UAV imageries from pipeline shake-table tests to generate seismic response data and assess seismic safety. The goal is to explore the UAV potential for seismic vibration monitoring including safety assessment of linear infrastructures by implementing several computer vision algorithms. The procedure is started by adopting the Maximally Stable Extremal Regions (MSER) method to extract co-variant regions that remain similar through a certain threshold of image series. The feature of interest is then detected, extracted, and matched using the Speeded-Up Robust Features (SURF) algorithm. It is assigned as artificial targets in this work that are distinctive from the surrounding environment to accelerate data processing. The Maximum Likelihood Estimation Sample Consensus (MLESAC) algorithm which is a generalization of the Random Sample Consensus (RANSAC) is applied for model fitting by maximizing the likelihood of the solution. The raw data are corrected using mathematical models and scaled to generate displacement data. Finally, a structural safety assessment is performed using several system identification models. These procedures are first validated using an aluminum bar placed on an actuator and tested under several harmonic tests followed by the large-scale pipeline shake-table test. The validation tests show a good agreement between the UAV data and reference data obtained using a stationary camera. The shake table tests also generate reasonable seismic performance and assess the pipeline seismic safety demonstrating the feasibility of the proposed procedure and the prospect of UAV-based SHM for linear infrastructure monitoring. Keywords: UAV-based SHM, computer vision, MSER, SURF, MLESAC, linear infrastructure, natural gas pipeline, seismic test, seismic performance.

Back to TopTop