No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Toward the balance between computational cost and model performance for the void detection of concrete-filled steel tubular structure using one-dimensional Mel-frequency cepstral coefficients and ensemble machine learning
... The incorporation of percussion method in CFST columns serves the purpose of ensuring optimal compaction and consolidation of the concrete. This technique employs percussion or vibration methods throughout the construction process to eliminate air voids, improve the bond between steel and concrete, and achieve a uniform density across the entire column [16,17]. To assess the concrete's density within the column tube, a sound-based analysis was employed, utilizing a specialized Number 3 steel hammer designed specifically for quality inspections. ...
... The incorporation of percussion method in CFST columns serves the purpose of ensuring optimal compaction and consolidation of the concrete. This technique employs percussion or vibration methods throughout the construction process to eliminate air voids, improve the bond between steel and concrete, and achieve a uniform density across the entire column [16,17]. To assess the concrete's density within the column tube, a sound-based analysis was employed, utilizing a specialized Number 3 steel hammer designed specifically for quality inspections. ...
Gradient Boosting Decision Tree (GBDT) is a popular machine learning algorithm , and has quite a few effective implementations such as XGBoost and pGBRT. Although many engineering optimizations have been adopted in these implementations , the efficiency and scalability are still unsatisfactory when the feature dimension is high and data size is large. A major reason is that for each feature, they need to scan all the data instances to estimate the information gain of all possible split points, which is very time consuming. To tackle this problem, we propose two novel techniques: Gradient-based One-Side Sampling (GOSS) and Exclusive Feature Bundling (EFB). With GOSS, we exclude a significant proportion of data instances with small gradients, and only use the rest to estimate the information gain. We prove that, since the data instances with larger gradients play a more important role in the computation of information gain, GOSS can obtain quite accurate estimation of the information gain with a much smaller data size. With EFB, we bundle mutually exclusive features (i.e., they rarely take nonzero values simultaneously), to reduce the number of features. We prove that finding the optimal bundling of exclusive features is NP-hard, but a greedy algorithm can achieve quite good approximation ratio (and thus can effectively reduce the number of features without hurting the accuracy of split point determination by much). We call our new GBDT implementation with GOSS and EFB LightGBM. Our experiments on multiple public datasets show that, LightGBM speeds up the training process of conventional GBDT by up to over 20 times while achieving almost the same accuracy.
The use of nano-materials to improve the engineering properties of different types of concrete composites including geopolymer concrete (GPC) has recently gained popularity. Numerous programs have been executed to investigate the mechanical properties of GPC. In general, compressive strength (CS) is an essential mechanical indicator for judging the quality of concrete. Traditional test methods for determining the CS of GPC are expensive, time-consuming and limiting due to the complicated interplay of a wide variety of mixing proportions and curing regimes. Therefore, in this study, artificial neural network (ANN), multi-expression programming, full quadratic, linear regression and M5P-tree machine learning techniques were used to predict the CS of GPC. In this instance, around 207 tested CS values were extracted from the literature and studied to promote the models. During the process of modeling, eleven effective variables were utilized as input model parameters, and one variable was utilized as an output. Four statistical indicators were used to judge how well the models worked, and the sensitivity analysis was carried out. According to the results, the ANN model calculated the CS of GPC with greater precision than the other models. On the other hand, the ratio of alkaline solution to the binder, molarity, NaOH content, curing temperature and concrete age have substantial effects on the CS of GPC.
In the present era of architecture, different cross-sectional shapes of structural concrete elements have been in use. However, this change in shape has a great effect on load-carrying capacity. To restore this capacity utilization of confinements of the column with elliptical sections has gained attention. This paper aimed to investigate the effect of elliptical sections of confined concrete with Carbon Fiber Reinforced Polymer (CFRP) and steel tube on axial load carrying capacity using Finite Element (FE) in Abaqus, analytical and Artificial Neural Networks (ANN) modeling. The study involved a column, 500 mm in height with three sets of aspect ratios, 1.0, 1.5, and 2.0. In each aspect ratio, three different layers of CFRP were used i.e., 0.167, 0.334, and 0.501 mm. Analytical results show that with the increase of aspect ratio from 1 to 2, there is a decrease in ultimate axial load of about 23.2% on average. In addition, the combined confining pressure of steel tube and CFRP increases with a decrease in dilation angle as several CFRP layers increases. The failure mode for the column with a large aspect ratio is a local buckling at its mid-height along the minor axis. The result showed a good correlation between FE and experimental results of ultimate stress and strains with the mean squared error of 2.27 and 0.001, respectively. Moreover, ANN and analytical models showed a delightful correlation of R2 of 0.97 for stress models and 0.88 for strain models, respectively. The elliptical concrete section of the column confined with steel tubes can be adopted for a new architectural type of construction, however, with more than three aspect ratios, the wrapping of the section with CFRP jackets is highly recommended.
Feature extraction and representation has significant impact on the performance of any machine learning method. Mel Frequency Cepstrum Coefficient (MFCC) is designed to model features of audio signal and is widely used in various fields. This paper aims to review the applications that the MFCC is used for in addition to some issues that facing the MFCC computation and its impact on the model performance. These issues include the use of MFCC for non-acoustic signals, adopting the MFCC alone or combining it with other features, the use of time series versus global representation of the MFCC, following the standard form of the MFCC computation versus modifying its parameters, and supplying the traditional machine learning methods versus the deep learning methods.
In this paper, a novel method to analyze the mechanical behavior of the coin-tap test using the principle of pounding mechanics is proposed. Based on the Hertzian contact theory, the linearized contact law is applied to simplify the mathematical model of the coin-tap test. By modeling the stiffness of the defect as an equivalent spring in a serial connection with the contact stiffness, the impact force and impact time can be theoretically derived. The derived expressions show that, the closer the defect is to the surface of the structure, the longer the impact duration time and the smaller the impact force will be. In addition, with the increase of defect radii, the peak force and duration of the impact will decrease and increase, respectively. Another innovation of this paper is that an equivalent impact stiffness (EIS)-based damage index (DI) is proposed to reveal the severity and location of the near-surface damages. Two validation experiments were carried out to validate the influences of the damage’s radii and under-surface depths. In the first experiment, the impact force and time of the central point for eight plates are measured to validate the accuracy of the proposed model. In the second experiment, 18 points on the horizontal centerline of eight specimens were tapped at an interval of 1 cm. The recorded pounding force and time were processed by EIS-based DI. Experimental results show that the EIS-based DI can effectively predict the position and severity of the defects and it holds the potential for practical application.
In recent decades, several studies have considered the use of plastic waste as a partial substitute for aggregate in green concrete. Such concrete has been limited to non-structural applications due to its low strength. This raises whether such concrete can be enhanced for use in some structural applications. This paper reports an attempt to develop a structural-grade concrete containing plastic waste aggregate with high proportions of substitution and confined with carbon fiber reinforced polymer (CFRP) fabrics. Experimental research was conducted involving the casting and testing 54 plain and confined concrete cylinders. A concrete mixture was designed in which the fine aggregate was partially replaced by polyethylene terephthalate (PET) waste plastic at ratios of 0%, 25%, and 50%, and with different w/c ratios of 0.40, 0.45, and 0.55. The results show that confinement has a substantial positive effect on the compressive behavior of PET concrete. The enhancement efficiency increases by 8–190%, with higher enhancement levels for higher substitution ratios. Adding one layer of CFRP fabric raises the ultimate strength of samples that have lost compressive strength to a level close to that of unconfined samples not containing PET. This confinement is accompanied by an increase in the slope of the stress-strain curve and greater axial and lateral strain values at failure. For the specimens confined by CFRP fabric, PET aggregate can be used as a partial substitute for sand at a replacement ratio of up to 50% by volume for structural applications. This paper also considers the ability of existing models to predict the strength of confined-PET concrete circular cross-sections by comparing model predictions with experimental results. The strength of confined PET concrete elements can’t be accurately predicted by any of the models that are already out there. It’s important to come up with a new model for these elements.
Concrete-filled steel tubes (CFSTs) are structural elements that, as a consequence of an incorrect elaboration, can exhibit internal defects that cannot be visualized, being usually air voids. In this work, the detection of internal damage in CFST samples elaborated with a percentage of contained air voids in concrete, was carried out by performing a complete ultrasound scan using an immersion tank. The analysis of the ultrasound signals shows the differences presented in the amplitude of the fundamental frequency of the signal, and in the Broadband Ultrasound Attenuation (BUA), in comparison with a sample without defects. The main contribution of this study is the application of the BUA technique in CFST samples for the location of air voids. The results present a linear relationship between BUA averages over the window of the CFSTs and the percentage of air voids contained (Pearson’s correlation coefficient r = 0.9873), the higher percentage of air voids, the higher values of BUA. The BUA algorithm could be applied effectively to distinguish areas with defects inside the CFSTs. Similar to the BUA results, the analysis in the frequency domain using the FFT and the STFT was sensitive in the detection of internal damage (Pearson’s correlation coefficient r = −0.9799, and r = −0.9672, respectively). The results establish an improvement in the evaluation of CFST elements for the detection of internal defects.
Internal cavities in structural elements, such as in concrete-filled steel tubes (CFSTs) of arch ribs reduce confinement, bearing capacity, and the durability of the arch bridges. Formation of internal cavities may be either materials related, which affects proper consolidation, or construction related, especially in terms of proper timing and delivery of concrete slurry to horizontal and inclined structural members. Nondestructive test methods such as Ultrasonics have been employed for detection of internal cavities. Despite their success, point-by-point inspection of structural members becomes time consuming and inefficient, especially when entire structures need to be inspected. Infrared thermography (IR) provides a more reasonable global approach for detection of anomalies. The accuracy and resolution of IR depends on the local ambient conditions affecting convection, and external thermal flux prior to detection by IR. The research described herein, pertains to the development of a hybrid thermographic approach by using Brillouin fiber optic sensors (BFOSs) for direct detection of thermal convection at the surfaces of steel tubes. Accurate quantitative detection of internal cavity locations and their dimensions required development of a specific machine learning based approach through which the distributed thermographic data was converted to a series of greyscale images. A Faster Region-based Convolutional Neural Network (Faster R-CNN) based on the improved VGG-16 (IVGG-16) network architecture was utilized for this purpose. The experimental work involved design and fabrication of a distributed temperature sensing sheet (DTSS) for simultaneous transmission of thermal energy into CFST specimens and acquisition of distributed thermographic data by the BFOS. By using the proposed approach, it was possible to detect the embedded internal cavities during the experiments.
The traditional monitoring methods can only give warnings for the bolts with severe looseness. However, it is essential for the safety of bolted joints to detect the looseness of bolts at the very early stage. To this end, in this paper, coda wave interferometry (CWI)-based high-resolution bolt preload monitoring using a single piezoceramic transducer is proposed. According to the CWI and acoustoelastic theories, a theoretical model is established and the linear relationship between the time shifts of coda waves and the preload variations of the bolt is derived. An experiment, in which a piezoceramic transducer simultaneously functions as the actuator and sensor, was carried out to verify the effectiveness of the proposed method. Three lead zirconium titanate transducers at different locations of a bolted specimen are tested. The experimental results show that the time shifts of coda waves increase linearly with the decrease of bolt preload and the detectable resolution of bolt preload (DRBP) is up to 0.326%. The DRBP value proves that the proposed technique can successfully monitor bolt looseness at the very early stage. In addition, a comparison study is carried out between the CWI-based method and the energy-based wavelet packet decomposition (WPD) method, and the result shows that the preload sensitivity of the CWI-based method is about six times higher than that of the WPD approach. Therefore, the CWI-based method is an effective way for the in situ monitoring of bolt looseness, especially in the embryonic stage.
The general approach to percussion-based monitoring of bolt preload is to train a classifier model to map the preloads and acoustic characteristics of percussion signals. However, the traditional percussion-based approach can only classify the preloads of bolts into certain ranges, and cannot accurately predict the exact bolt preload. This paper proposes an approach to estimate bolt preload using a convolutional neural network (CNN). The frequency contents of the percussion signals are analysed with the Fast Fourier Transform (FFT), and the magnitudes of signals in different frequencies ranges are reconstructed into a matrix, which can be treated as an image. Each image is labelled with the corresponding bolt preload. Then the labelled images are used to train the CNN model, and the trained model is used to detect the actual preload of a selected bolt. The proposed model was experimentally verified using a bolted steel plate. The results show that the proposed model can accurately predict the preload values outside the range of training samples with an accuracy over 95%. Meanwhile, compared with the predicted effect of the traditional regression models, the predicted validity of the proposed model is further verified.
This paper proposes a new concept, named the Detectable Resolution of Bolt Pre-load (DRBP), by using the coda wave interferometry (CWI) to quantitatively measure the pre-load looseness at a high resolution. Due to its characteristics of roughness, irregularity, and randomly distributed asperities, the contact surface of the bolted components can function as a natural interferometer to scatter the propagation waves. The multiply-scattered coda waves can amplify the slight changes in the travel path and show the visible perturbation in the time domain. By calculating the time-shifted correlation coefficient of coda waves before and after the slight pre-load looseness, the tiny pre-load changes can be clearly revealed. To evaluate the feasibility of the proposed method, a theoretical model considering the time shifts of coda waves and the variations of pre-load is established. Based on the acoustoelastic effect and the wave path summation theory of coda wave interferometry, the model shows that the time shifts of coda waves change linearly with the variations of pre-load. Verification experiments are conducted, and the results show that the R-square values of the fitting curves are larger than 0.9216. In addition, the proposed approach has the feature of high resolution. The ultimate pre-load resolution of the proposed approach is 0.331%, that is, when the variation of pre-load is larger than 0.331%, it can be detected. Therefore, theoretical analysis and experimental results prove that the CWI-based pre-load detection approach holds great potential for the detection of bolt pre-load looseness, especially during the initial stage.
A novel composite structural member, UHPC-encased concrete-filled steel tubular (UC-CFST) column, is proposed in this study to address the shortcomings of the ordinary concrete-encased CFST (OC-CFST) columns. A total of 22 specimens subjected to axial compression loading were tested, including nine UC-CFST stub columns, nine OC-CFST stub columns, and four CFST stub columns. The experimental results showed that the encasing UHPC of UC-CFST columns was less damaged than the encasing ordinary concrete of OC-CFST columns and did not spall when the UHPC was crushed. The UC-CFST stub columns reached their ultimate load-bearing capacity when the encasing UHPC attained the peak compressive strain, which was much greater than that of the OC-CFST stub columns. Compared to the OC-CFST stub columns, the ultimate load-bearing capacity of UC-CFST stub columns improved markedly by 75%–289% and the compressive stiffness increased by 22%–49%, while the ductility decreased by about 50% on average. Furthermore, increasing the area ratio of the encased CFST in the UC-CFST stub columns was unfavorable for the compressive stiffness and load-ultimate bearing capacity. Finally, by comparing the different calculation methods for ultimate load-bearing capacity of OC-CFST columns recommended by Chinese standard CECS 188, American standard AISC 360, Japanese standard AIJ-SRC, and European standard Eurocode 4, it was found that the calculation methods applicable to OC-CFST columns would underestimate the ultimate load-bearing capacity of UC-CFST columns to varying degrees. Thus, a modified calculation formula based on the CECS 188 standard was proposed for accurate prediction of the ultimate load-bearing capacity of the UC-CFST stub columns.
A real-time and durable system for scour process monitoring with sufficient spatial precision is in pressing need for bridge safety management. In light of this, an innovative bundled time domain reflectometry sensing cable was recently proposed to enhance the time domain reflectometry technique for scour monitoring. However, current development only dealt with the construction of bundled sensing cable and the corresponding new data reduction method. Before it can be put into practical use, issues related to the effect of hydrological conditions, long-distance measurement, and actual field implementation are yet to be investigated. This study used both numerical simulations and laboratory experiments to examine the time domain reflectometry signals in response to both scour and deposition in different water-level conditions. As a result, the first guideline of waveform classification and interpretation is newly proposed to validly determine scour depth under various field conditions. Since field measurements often come with significant signal attenuation from resistance loss of long cable and dielectric and conductive loss in the sensing section, numerical simulations and a series of full-scale experiments were also conducted to assess the time domain reflectometry scour measurement range. The maximum measurement range of the latest time domain reflectometry scour sensing cable was found to be about 6 m. Within this range, the maximum error of scour estimation is within 0.2 m. Considering the new findings, a field time domain reflectometry scour monitoring system using the bundled time domain reflectometry sensing cable was designed accordingly and implemented at a bridge for the first time. The monitoring system successfully captured the scour process during a storm event and revealed some practical issues for future improvement as well.
The void defects significantly threaten the integrity of concrete-filled steel tubular (CFST) structure. Along with detecting the presence of subsurface voids, the knowledge of the geometry of voids can provide meaningful information to determine the overall structural health. This study develops a method to determine the depths of subsurface voids by integrating the percussion technique with a machine learning algorithm. A decision tree model is trained to detect different depths of subsurface voids in a CFST specimen. The percussed sounds from areas with and without subsurface voids were analyzed by using the power spectrum density (PSD). The process was repeated 100 times, and the average correction ratio is up to 96.33%. The results showed that the proposed approach has great potential in subsurface void inspection and evaluation for CFST structures.
Corrosion of steel bars leads to significant structural deteriorations in reinforced concrete structures, increasing their maintenance costs and shortening their service life. Fiber Reinforced Polymer (FRP) bars, as an internal reinforcing material instead of steel bars, are used in concrete structures owing to its high tensile strength and corrosion resistance. However, the structures of FRP reinforced concrete bending components have the large deflection and the lower post-cracking bending stiffness. In addition, it is difficult to evaluate the bending stiffness of in service FRP reinforced concrete beam by using the traditional monitoring method. This paper proposes a novel approach to real-time monitoring of the bending stiffness of FRP reinforced concrete beams using piezoceramic transducers enabled stress wave propagation. In this approach, several piezoceramic smart aggregate (SA) transducers are bonded on the side-surface of a concrete beam reinforced with Basalt-FRP (BFRP) bars to evaluate the bending stiffness based on stress wave propagation. A piezoceramic SA transducers based bending stiffness index (Piezo-BSI) is proposed to quantify the bending stiffness levels of BFRP reinforced concrete beams. The results show that the bending stiffness of BFRP reinforced concrete beams can be effectively evaluated by using SA transducers. The proposed Piezo-BSI values agree well with the actual bending stiffness index. This indicates that the Piezo-BSI values can accurately quantify and effectively reflect the actual bending stiffness levels of concrete beams reinforced with BFRP bars.
This article presents a novel framework for monitoring and evaluation of a population of bridges using smartphones in a large number of moving vehicles as mobile sensors. Within this framework, a damage detection methodology based on Mel-frequency cepstral coefficients and Kullback–Leibler divergence is developed. For this method, Mel-frequency cepstral coefficients of the vibration data collected from smartphones in vehicles crossing bridges are first extracted as features. Then, Kullback–Leibler divergence is used to compare the distributions of features. The damage in a bridge can be identified by quantifying the difference of the distributions obtained for the same bridge. Both numerical and lab experiments are conducted to verify the proposed framework and methodology. In lab experiments, a smartphone and two wireless accelerometers are used for data collection. From our results, it is concluded that the damage existence can be successfully identified using smartphones in a large number of vehicles. Also, it is observed that there is a significant correlation between the magnitude of the damage features and the severity of damage. The results show that the method has the potential to monitor a population of bridges simultaneously and in almost real time.
Predictive analytics play an important role in the management of decentralised energy systems. Prediction models of uncontrolled variables (e.g., renewable energy sources generation, building energy consumption) are required to optimally manage electrical and thermal grids, making informed decisions and for fault detection and diagnosis. The paper presents a comprehensive study to compare tree-based ensemble machine learning models (random forest – RF and extra trees – ET), decision trees (DT) and support vector regression (SVR) to predict the useful hourly energy from a solar thermal collector system. The developed models were compared based on their generalisation ability (stability), accuracy and computational cost. It was found that RF and ET have comparable predictive power and are equally applicable for predicting useful solar thermal energy (USTE), with root mean square error (RMSE) values of 6.86 and 7.12 on the testing dataset, respectively. Amongst the studied algorithms, DT is the most computationally efficient method as it requires significantly less training time. However, it is less accurate (RMSE = 8.76) than RF and ET. The training time of SVR was 1287.80 ms, which was approximately three times higher than the ET training time.
The impact-echo (IE) method is a popular non-destructive testing (NDT) technique widely used for measuring the thickness of plate-like structures and for detecting certain defects inside concrete elements or structures. However, the IE method is not effective for full condition assessment (i.e., defect detection, defect diagnosis, defect sizing and location), because the simple frequency spectrum analysis involved in the existing IE method is not sufficient to capture the IE signal patterns associated with different conditions. In this paper, we attempt to enhance the IE technique and enable it for full condition assessment of concrete elements by introducing advanced machine learning techniques for performing comprehensive analysis and pattern recognition of IE signals. Specifically, we use wavelet decomposition for extracting signatures or features out of the raw IE signals and apply extreme learning machine, one of the recently developed machine learning techniques, as classification models for full condition assessment. To validate the capabilities of the proposed method, we build a number of specimens with various types, sizes, and locations of defects and perform IE testing on these specimens in a lab environment. Based on analysis of the collected IE signals using the proposed machine learning based IE method, we demonstrate that the proposed method is effective in performing full condition assessment of concrete elements or structures.
Recently, in the field of structural health monitoring, the detection of bolted connection looseness through percussion-based method and machine learning technology has received much attention due to the advantages of removing the requirement of sensor installation and potential for automation. However, there are few such research which are performed in the underwater environment. The paper proposes a new method, Feature-reduced Multiple Random Convolution Kernel Transform (FM-ROCKET), to identify the looseness level of the underwater bolted connections based on the percussion-induced sound (audio signal). By integrating deep learning (DL) and shallow learning, the FM-ROCKET model uses the 1D convolutional layer (a DL method) to extract features from the percussion-induced audio signal and adopts the rigid classifier (linear classifier, a shallow learning method) to classify the features. Five different preload levels of the bolted flange are considered. A hammer is utilized to tap the flange surface and the continuous percussion-induced audio signal is collected by a smartphone in an underwater environment. After the audio signal segmentation, single-hit audio signals are fed into the FM-ROCKET model. To verify the effectiveness of the proposed method, three case studies are conducted on two flanges. In case study I, the proposed method slightly outperforms other DL-based methods under different training/test splitting ratios. In case studies II and III, the proposed method is far more effective than other DL-based methods on independent and different test sets. The results demonstrate the superiority of the FM-ROCKET model in the underwater detection of bolted flange looseness. To the best of our knowledge, this article is the first attempt to address the detection of bolted flange looseness in the underwater environment by combining percussion-based method, DL, and shallow learning.
Concrete-filled steel tubular (CFST) structures have been widely used to provide high load-bearing capacity. Percussions are expected to provide simple and effective means of detecting subsurface voids of CFST structures, caused by poor grouting compactness and harsh operating conditions. Existing percussion data processing methods consist of two steps to deal with signals with high sampling rates: heuristic feature extraction and classification. High-dimensional feature representations tend to be selected to achieve visual interpretability and performance, increasing the complexity and computational cost of the entire process. This research aims at developing a computationally efficient end-to-end framework for the percussion-based void detection of CFST structures suitable for mobile computing devices available in the field, while maintaining the performance and a certain degree of visual interpretability. The proposed approach is based on the dense learned features; First, a classifier that processes raw percussion sound signals is developed using dilated convolutions and global max-pooling. The class activation mapping (CAM) technique is then applied to convert the classifier into the dense feature extractor. Feature visualization for the experimental data shows that (i) the classifier listens to the "echo" of the percussion sound carefully, rather than the impact sound itself, and (ii) the classification results depend on both the vibration frequency and its damping ratio. Finally, the feature extractor is extended to quasi real-time void detection and classification, where the processing time duration for a single 0.1-second (50,000-point) percussion signal is 8.28 ms without leveraging advanced computing capabilities of graphics processing units, and the average testing accuracy for the fivefold cross validation is 99.81%. The results demonstrate the potential of the proposed approach for improving the efficiency of inspections.
To study the effect of slenderness ratio on the ultimate bearing capacity of UHPC-encased concrete-filled steel tubular (UC-CFST) columns, a total of nine specimens under axial compression were tested. A finite element (FE) model, considering the effects of material nonlinearity and interaction between concrete and steel tube, was developed. The full-range analysis was carried out on the typical load-deformation response of the UC-CFST medium-long columns. The experimental and numerical analysis results showed that with the increase in slenderness ratio, the deformation capacity of the UC-CFST columns increased, the ultimate bearing capacity decreased, and the failure mode changed from exceedance of compressive strength to loss of stability. The internal forces and stress distributions at the mid-height section of the columns at different stages were also investigated. Finally, the applicability of various methods for calculating the ultimate bearing capacity of the ordinary concrete-encased CFST (OC-CFST) columns to the UC-CFST columns was discussed, including the methods recommended by Chinese standard T/CECS 188-2019, European standard Eurocode4, Australian/New Zealand joint standard AS/NZS 5100.6:2017, and American standard ANSI/AISC 360-16. It was found that the method recommended in the AS/NZS 5100.6:2017 was suitable for calculating the ultimate bearing capacity of the UC-CFST medium-long columns.
Concrete has been widely used in underwater structures. However, accidental voids near the concrete protective layer may weaken the bearing capacity of the structures and can even lead to severe disasters. Therefore, it is necessary to develop an efficient and convenient method for quantitatively identifying voids in concrete and guaranteeing structural safety. A percussion–acoustic method is commonly used to identify voids in concrete exposed to air, owing to its high efficiency and low‐cost. Nevertheless, its feasibility for underwater concrete remains questionable, owing to the complexity of the underwater environment. To address this limitation, this study conducted experimental and theoretical studies by using a percussion–acoustic‐based method to identify voids in underwater concrete while considering the fluid–structure coupling effect. The results demonstrate that the frequency characteristics of percussive sound pressure underwater are significantly different from those in air and are determined by the void scopes and depths. Accordingly, this study provides a potentially applicable method for evaluating the voids in underwater concrete.
Indirect bridge health monitoring (iBHM) has gained significant attention in recent years as an alternative to direct bridge monitoring techniques. iBHM leverages the use of a vehicle traveling over a bridge as a sensing device as well as an actuator. The resulting data acquired from the vehicle contains dynamic information about the bridge, vehicle suspension, engine noise, ambient traffic noise, road roughness, etc. Due to the presence of sensor-vehicle-bridge interactions in vehicle measurement, a powerful decoupling framework of system identification is required to extract useful information about the bridge and delineate the effect of vehicle and driving frequencies. This paper proposes a hybrid time-frequency method capable of decoupling vehicle-bridge interactions of vehicle measurement and performing a robust bridge modal identification under various operational challenges. In the proposed research, wavelet packet transformation (WPT) is first used to decompose the vehicle measurement into various WPT coefficients. Then, the modal frequencies of the bridge are extracted from the resulting WPT coefficients using Synchro-Extracting Transform (SET) and are separated from the vehicle and driving frequencies. The proposed method is validated using numerical and laboratory studies as well as a full-scale study comprising of a 220 m long box-girder bridge. The results provide strong evidence that the proposed hybrid time-frequency method can serve as a promising iBHM method.
Detecting the initial damage of the engine precisely is conducive to finding out the failure symptom timely and ensuring the reliable operation of the engine for avoiding malignant accidents. This article proposes a single-sensor acoustic emission (AE) diagnosis method based on deep learning (DL) for structural health monitoring (SHM) and initial damage intelligent identification of the engine. Firstly, the time-frequency domain features of AE signals collected from the engine test bench under different initial damages are extracted by continuous wavelet transform (CWT), and two-dimensional time-frequency images are constructed using the obtained results. After that, the proposed deep convolutional neural networks (DCNN) is employed to directly extract damage features from the converted two-dimensional time-frequency images via a multi-layer fusion and perform initial damage intelligent identification. In addition, the white Gaussian noise with different signal-to-noise ratio (SNR) is added to the raw AE signal, and the robustness of the proposed DCNN method is verified. Finally, to evaluate the performance of the proposed DCNN, the constructed test accuracy Euclidean distance and other indexes including Kappa, microF1 and macroF1 are introduced to compare with other commonly machine learning methods. The results demonstrated the effectiveness of the proposed DCNN model, which lays the foundation for the application of the single-sensor AE diagnosis method based on DCNN in SHM and initial damage intelligent identification of the engine.
This paper presents the behavior of recycled aggregate concrete-filled steel tubular columns under combined compression and shear load. A total of 30 recycled aggregate concrete-filled steel tubular specimens were tested, with their failure modes and shear capacities investigated. The main parameters considered include the cross-section type, the axial load ratio, the shear span-to-depth ratio, the steel ratio and the replacement rate of recycled coarse aggregates. A finite element analysis model is established, which is verified and then used to conduct analysis on the shear performance in terms of the failure modes, the full-range load-deflection relations, the stress distributions and the load transfer mechanism. The results show that both the axial load ratio and the shear span-to-depth ratio have influence on the failure modes. A strut-tie mechanism model is proposed to reveal these influences. Four failure modes are identified, while recycled aggregate concrete-filled steel tubular columns tend to fail in a shear manner with a smaller shear span-to-depth ratio and a higher axial load level. Finally, an existing simplified model is assessed to predict the shear capacity of recycled aggregate concrete-filled steel tubular columns under combined compression and shear load.
As one of the most common coupling elements in infrastructures, bolted joints play an important part in ensuring the integrity and safety of the whole system, whose failure may cause disastrous consequences. In recent years, precise detection and evaluation of bolt looseness have attracted numerous researchers' interest. However, the reliability of existing methods cannot be well guaranteed in long-term field detection, and real-time feedback is rather costly. This paper proposes a novel bolt looseness detection method based on audio recognition and deep learning. Firstly, a percussion experiment was designed to collect audio signals of bolts at different torque levels. Then, the time-domain bolt percussion signals were converted into Mel-frequency spectrograms, and the convolutional neural network (CNN) was adopted to mining deep information from the images for classification. To further verify the effect of different initial prestress levels on the vibration frequency of the bolted joint, a numerical study was conducted with the consideration of three different prestress levels. The results reveal that the proposed method has a high recognition accuracy in identifying bolt looseness conditions. Additionally, an iOS APP of acoustic vibration was established for real application. The prerecorded and untrained percussion audio was used to simulate the real-time bolt looseness detection, which shows its potential in real future applications.
A new structural type of seawater and sea sand concrete (SWSSC)-filled fibre-reinforced polymer (FRP) and steel wire mesh composite tube (SFWCT) was proposed. The steel wire mesh, working as an intermediate layer, was embedded in two layers of FRP. Due to the high corrosion resistance of FRP, the steel wire mesh could maintain a long-term stable working state while the core concrete was effectively confined. Three series of specimens were tested under compressive loading. The test parameters mainly included the number of layers of steel wire mesh, the FRP type and the FRP thickness. The results show that the time to failure was significantly reduced with the new design, as the steel wire mesh could effectively decrease the degree of damage in the specimens and prevent the specimens from immediately losing their bearing capability due to the sudden fracture of the FRP. The accuracies of relevant models were evaluated with the test data, and the model with the best performance was suggested. Moreover, the full stress–strain curves of SFWCTs under compression were simulated, and the calculated results were satisfactory.
Bolted connection functions to fasten and secure parts together for engineering structures. Newly developed percussion-based approaches have been proven as a fast and effective tool for bolt looseness identification; however, most of the existing studies use machine learning assisted approaches to classify percussion sounds and predict looseness conditions without investigating the relationship between percussion sounds and bolt vibrations. This paper presents a conceptual research to utilize bolt vibration signal to reconstruct percussion sounds, which significantly improves the validity and accuracy of bolt looseness identification. In the experimental study, a laser Doppler vibrometry was used to capture vibrational information of the test bolt; meanwhile, percussion sounds were collected by microphones. The relationship between sound and vibration signals was investigated using wavelet packet decomposition and correlation analysis. A new set of sounds were reconstructed by combination of the sound packets which showed the strongest correlation with the vibration signal. The reconstructed sound database was then transformed into spectrograms and trained by a two-dimensional convolution neural network to identify bolt looseness conditions.
The induction of decision trees is a widely-used approach to build classification models that guarantee high performance and expressiveness. Since a recursive-partitioning strategy guided for some splitting criterion is commonly used to induce these classifiers, overfitting, attribute selection bias, and instability to small training set changes are well-known problems in them. Other approaches, such as incremental induction, classifier ensembles, and the global search in the decision-tree-space, have been implemented to overcome these problems. In particular, metaheuristics such as simulated annealing, genetic algorithms, genetic programming, and ant colony optimization have been used to induce compact and accurate decision trees. This paper presents a state-of-the-art review of the use of single-solution-based metaheuristics and swarm and evolutionary computation algorithms to build decision trees as classification models. We outline the decision-tree-induction process components and detail the existing literature studies on metaheuristic-based approaches to building these classifiers. Several timelines showing the chronological order in which these approaches were introduced in the literature are included. A summary analysis of these studies is also conducted, focusing on their internal components and experimental studies. This work provides a useful reference point for future research in this field.
Concrete-filled steel tubular (CFST) column may be undergone coupled compression and torsion during its service life, and therefore, six CFST specimens were tested to analyze the torsional behaviour. The failure modes, torsion moment (T) versus torsional angle (θ) relationships and torsion moment-strain curves derived from the testing results for CFST specimens under different torsion conditions were compared and analyzed. Meanwhile, the numerical model of CFST members was developed via the software ABAQUS to further analyze the torsional mechanism. Moreover, the comparison of torsional performance between CFST column and steel-reinforced concrete-filled steel tubular (SRCFST) column was conducted. The results indicated that the T-θ curve of each specimen showed slight difference in the initial stage under diverse loading conditions; however, the specimens under torsion after compression exhibited highest ultimate torsional capacity, followed by the specimens under simultaneous compression-torsion and pure torsion. The T-θ curve of SRCFST columns with I-section steel or cruciform steel was similar to that of CFST columns, but the T-θ curve of SRCFST columns increased obviously in the later stage due to the inserted steel tube. In addition, the validity of the numerical model was verified against experiment. Then, the interaction force between steel tube and concrete and the internal forces of the characteristic points of the T-θ curves under different loading conditions and axial load levels were compared to understand the stress mechanism. Finally, the proposed calculation method for torsional capacity of CFST members under coupled compression-torsion could provide a more accurate prediction.
Currently, the steel–concrete composite structures (SCCSs) have been widely applied as primary load-bearing members owing to their excellent structural behavior. However, due to the comprehensive influence of cyclic loading, periodical temperature, unscientific pouring process and severe corrosive service environment, interfacial imperfections including bond-slip and debonding defects occur, resulting in the deterioration of serviceability and durability. Therefore, it is urgent to develop effective nondestructive testing (NDT) methods for SCCSs to guarantee their mechanical performance and structural safety. In this study, a state-of-the-art review on existing interfacial imperfection detection technologies for SCCSs using NDT techniques is conducted. The typical structural type of SCCSs and characteristics of interfacial imperfections are reviewed first. The basic detection principle and mechanism of direct measuring and empirical methods, smart materials and sensors-based NDT technology, as well as corresponding typical applications, are comparatively discussed in detail. More particularly, the stress wave-based NDT testing is separately discussed, owing to its extensive application in recent years. In addition, the advantages and disadvantages, similarities and differences between various NDT methods are systematically compared and analyzed in depth. The research findings can provide constructive guidance for the scientific optimization of future experimental study and essential guidance to practical NDT testing on SCCSs.
In order to detect interfacial imperfections for steel–concrete composite structures (SCCSs), various advanced non-destructive testing (NDT) methods have been proposed in recent years. Wave-based NDT is known for their unique advantages, including precise detection mechanisms, uncomplicated acquisition system and reliable measurements, and have been extensively applied in NDT of SCCSs. This manuscript is a comprehensive review covering literature that describes research in damage detection for SCCSs. The review includes a detailed analysis of the state-of-the-art research in the experimentation, multiscale simulation and multi-physics coupling analysis of wave method-based NDT techniques for SCCS. First, the practical application of SCCSs and defects typically observed in SCCSs are reviewed. Wave propagation analysis using multiscale models established with the random aggregate method is compared against macro-level numerical simulation results. Existing multi-physics coupling analysis of piezoelectric lead zirconate titanate (PZT)-based stress wave methods are classified and summarized. This area of research is forecasted to enhance the effectiveness and the multi-functionality of future wave method-based NDT techniques for SCCSs. This review offers guidance and references to numerical investigation on detection mechanism behind stress wave-based monitoring and test design of wave-based NDT experiments.
Concrete-filled steel tubular (CFST) columns have been used in the construction of modern structures such as high-rise buildings and bridges as well as infrastructures as they provide better structural performance than conventional reinforced concrete or steel members. Different shapes of CFST columns may be needed to satisfy the architectural and aesthetic criteria. In the study, three dimensional FE simulations of circular, square, hexagonal , and octagonal CFST stub columns under axial compression were developed and verified through the experimental test data from the perspectives of full load-displacement histories, ultimate axial strengths, and failure modes. The verified FE models were used to investigate and compare the structural performance of CFST columns with different cross-section shapes by evaluating the overall load-deformation curves, interaction stress-deformation responses, and composite actions of the column. The extent of the ultimate-axial-strength enhancement due to enhanced steel yield strength and concrete compressive strength was evaluated through the parametric studies. At last, the accuracy of available design models in predicting the ultimate axial strengths of CFST columns were investigated. Research results showed that the behaviors of hexagonal and octagonal CFST columns were generally similar to that of the square CFST column as their overall structural performance was relatively improved.
In proportion to the immense construction of spatial structures is the emergence of catastrophes related to structural damages (e.g. loose connections), thus rendering personal injury and property loss. It is therefore essential to detect spatial bolt looseness. Current methods for detecting spatial bolt looseness mostly focus on contact-type measurement, which may not be practical in some cases. Thus, inspired by the sound-based human diagnostic approach, we develop a novel percussion method using the Mel-frequency cepstral coefficient and the memory-augmented neural network in this article. In comparison with current investigations, the main contribution of this article is the detection of multi-bolt looseness for the first time with higher accuracy than prior methods. In particular, in terms of new data obtained via similar joints, the memory-augmented neural network can help avoid inefficient relearn and assimilate new data to provide accurate prediction with only a few data samples, which effectively improves the robustness of detection. Furthermore, percussion was implemented with a robotic arm instead of manual operation, which preliminarily explores the potential of implementing automation applications in real industries. Finally, experimental results demonstrate the effectiveness of the proposed method, which can guide future development of cyber-physics systems for structural health detection.
Language discrimination among similar languages, varieties, and dialects is a challenging natural language processing task. The traditional text-driven focus leads to poor results. In this article, we explore the effectiveness of speech-driven features toward language discrimination among Chinese dialects. First, we systematically explore the appropriateness of speech-driven MFCC features toward CNN-based language discrimination. Then, we design an end-to-end speech recognition model based on HMM-DNN to predict Chinese dialect words. We adopt attention mechanism to extract the discriminative words related to different Chinese dialects. Finally, through a CNN, we combine the word-level embedding and the MFCC-based features. Evaluation of two benchmark Chinese dialect corpora shows the appropriateness and effectiveness of the proposed speech-driven approach to fine-grained Chinese dialect discrimination compared to the state-of-the-art methods.
Across multiple construction processes, the cup-lock scaffolding systems have been widely applied as temporary facilities, while several catastrophes caused by scaffolding collapse have been reported. Therefore, in this paper, we conduct an exploratory study to attempt to research one issue that can affect the stability of scaffolding systems, namely the looseness of the cup-lock joint. In this paper, to detect looseness of cup-lock scaffold, we develop a new percussion-based method to avoid current structural health monitoring (SHM) methods that depend on constant contact between structures and sensors. Particularly, inspired by the rapid development of automatic speech recognition (ASR), we propose a convolutional bi-directional long short-term memory (CBLSTM) model to classify the Mel frequency cepstral coefficient (MFCC) features extracted from percussion-induced sound signals. To the best of our knowledge, this is the first application of ASR technique in looseness detection of cup-lock scaffold. The working mechanism of CBLSTM is given as follows: a convolutional neural network (CNN) is used to craft characteristics from MFCC features, and a bi-directional long short-term memory architecture (BLSTM) can improve classification accuracy by assimilating the learned CNN features. Finally, a laboratory experiment is conducted to verify the effectiveness of the proposed method, and we demonstrate that CBLSTM outperforms the CNN and BLSTM in classifying the MFCC features. Overall, the percussion-based method proposed in this paper can provide a new direction for the investigation, particularly health monitoring, on the cup-lock scaffolding system.
Deposits prevention and removal in pipeline has great importance to ensure pipeline operation. Selecting a suitable removal time based on the composition and mass of the deposits not only reduces cost but also improves efficiency. In this article, we develop a new non-destructive approach using the percussion method and voice recognition with support vector machine to detect the sandy deposits in the steel pipeline. Particularly, as the mass of sandy deposits in the pipeline changes, the impact-induced sound signals will be different. A commonly used voice recognition feature, Mel-Frequency Cepstrum Coefficients, which represent the result of a cosine transform of the real logarithm of the short-term energy spectrum on a Mel-frequency scale, is adopted in this research and Mel-Frequency Cepstrum Coefficients are extracted from the obtained sound signals. A support vector machine model was employed to identify the sandy deposits with different mass values by classifying energy summation and Mel-Frequency Cepstrum Coefficients. In addition, the classification accuracies of energy summation and Mel-Frequency Cepstrum Coefficients are compared. The experimental results demonstrated that Mel-Frequency Cepstrum Coefficients perform better in pipeline deposits detection and have great potential in acoustic recognition for structural health monitoring. In addition, the proposed Mel-Frequency Cepstrum Coefficients–based pipeline deposits monitoring model can estimate the deposits in the pipeline with high accuracy. Moreover, compared with current non-destructive deposits detection approaches, the percussion method is easy to implement. With the rapid development of artificial intelligence and acoustic recognition, the proposed method can realize higher accuracy and higher speed in the detection of pipeline deposits, and has great application potential in the future. In addition, the proposed percussion method can enable robotic-based inspection for large-scale implementation.
A novel type of seawater and sea sand concrete (SSC)-filled FRP-carbon steel composite tube (SFSCT) column composed of FRP-carbon steel composite tubes and core-filled SSC is proposed. The FRP carbon steel composite tube is made by wrapping FRP sheets around the inner and outer walls of steel tubes to isolate the transport of chloride ions from the external corrosive environment and the internal SSC, respectively. A total of 36 SFSCT columns were tested under axial compression to investigate the axial compressive performance. The test parameters include the FRP thickness, FRP type and steel tube thickness. The test results show that internal FRPs and external FRPs can work together and effectively improve the bearing and deformation capacities of the structure under axial compression. Compared with concrete-filled steel tube (CFST) columns, with different numbers of FRP wrapping layers, the strength of the SFSCT columns can be increased by 11.7 - 66.5%. The confinement effect of CFRP is better than that of BFRP. Compared with the steel tube thickness, the FRP type and FRP thickness have a more significant influence on the stress-strain behaviour of SFSCT columns.
Mode identification in civil engineering uses ambient excitation response, where the excitation is assumed to be stationary white noise. However, practical engineering conditions cannot satisfy this assumption of ideal excitation. This paper proposes an innovative method based on eigensystem realization algorithm and a proposed virtual frequency response function. The formulation of the virtual frequency response function is derived from the ratio of the cross‐power and autopower spectral density functions of the measurement signals. The impulse responses are calculated by inverse Fourier transform of the product of the virtual frequency response function and the frequency response function of the reference point. After obtaining impulse responses, eigensystem realization algorithm is then performed to identify structural modes. The proposed virtual frequency response function and eigensystem realization algorithm are validated by a numerical example. The results show that the virtual frequency response function and eigensystem realization algorithm can give much a better impulse response than natural excitation technique and identify more precise mode parameters. Finally, the proposed method is applied to a practical engineering bridge, where the results by the virtual frequency response function and eigensystem realization algorithm can also give better results than those given by natural excitation technique and eigensystem realization algorithm.
Decision tree classification has become a prevailing technique for online diagnosis services. By outsourcing computation intensive tasks to a cloud server, cloud-assisted online diagnosis services are better ways for cases that the storage and computation requirements exceed the capability of medical institutions. With privacy concerns as well as intellectual property protection issues, the valuable diagnosis classifier and the sensitive user data should be protected against the cloud server. In this paper, we identify a work-flow for cloud-assisted online diagnosis services. We propose an efficient and secure decision tree classification scheme in the proposed work-flow. Specifically, the medical institution transforms a locally pre-trained decision tree classifier to a decision table, and later uses searchable symmetric encryption to encrypt the decision table. Then, the encrypted table is outsourced to the cloud server, and a user can submit encrypted physiological features to the cloud server and obtain an encrypted diagnosis prediction back. We provide formal security proofs to demonstrate that our scheme protects the confidentiality of the decision tree classifier and the user's data. The performance analysis shows that our scheme achieves faster-than-linear classification speed. Experimental evaluations show that our scheme requires several micro-seconds to process a diagnosis request in the tested datasets.
Special-shaped CFST columns are becoming increasingly attractive as alternative solutions to engineering design. Three-dimensional FE models are developed and verified against experimental results in terms of failure modes, load-deformation curves and ultimate loads, where circular, triangular, Fan-shaped, D-shaped, 1/4 circular and semi-circular sections are considered. In light of the FE simulations, the composite actions between the special-shaped steel tubes and concrete cores have been investigated through load-deformation and interaction stress-deformation histories. Possible parameters affecting specimens loading behaviors have been studied. The studies generally show that the failure modes, composite behaviors and load-deformation histories of the axially loaded special-shaped CFST stub columns are similar to those of SHS/RHS specimens.
In practical engineering, power spectral density is always used to identify structural dynamic properties by picking peak points. However, the peaks in the spectral curve are not easy to be determined sometimes. A lot of peaks in the curve would cause some peaks close, which induces the closely spaced modes. The damping ratio is another reason to generate closely spaced modes, because the damping can make some peaks merge together. This paper proposes an innovative method to identify the closely spaced modes. First, the formulation is derived to reveal the components of the closely spaced modes in the power spectral density based on frequency domain decomposition. Then singular vector comparison is presented to determine whether there are closely spaced modes or not, where an angle criterion between two vectors is proposed. Finally, a numerical example is used to validate the effectiveness of the proposed method. The results show that the angle criterion can find the independent and dependent modes. For the close peak points in the spectral curve, the angle can determine that the points are single or closely spaced modes. Therefore, the proposed method to identify the closely spaced modes is efficient.
Ensemble methods are considered the state‐of‐the art solution for many machine learning challenges. Such methods improve the predictive performance of a single model by training multiple models and combining their predictions. This paper introduce the concept of ensemble learning, reviews traditional, novel and state‐of‐the‐art ensemble methods and discusses current challenges and trends in the field.
This article is categorized under:
• Algorithmic Development > Model Combining
• Technologies > Machine Learning
• Technologies > Classification
The use of high strength concrete and steel have significant advantages for composite members subject to significant compression as in the cases of high-rise buildings. Current design codes place limits on the strengths of steel and concrete due to limited test data and experience on the behaviour of composite members with the high strength materials. To extend their applications, a comprehensive experimental program has been carried out to investigate the behaviour of concrete filled steel tubes (CFSTs) with high- and ultra-high- strength materials at ambient temperature. This article presented some new findings on the axial performance of 56 short CFSTs. High tensile steel with yield strength up to 780 MPa and ultra-high strength concrete with compressive cylinder strength up to 190 MPa were used to prepare the CFST test specimens. The key issue is to clarify if the plastic cross-sectional resistance could be used at ultimate limit state as for CFSTs with the normal strength materials. To address this, experimental and analytical methods were adopted where the test results were compared with the predictions by various design codes world widely, and design recommendations were therefore proposed so that the prediction methods could be safely extended to the short CFSTs with the high- and ultra-high- strength materials.
This paper numerically studied the collapse capacity of high-rise steel moment-resisting frames (SMRFs) using various width-to-thickness members subjected to successive earthquakes. It was found that the long-period component of earthquakes obviously correlates with the first-mode period of high-rises controlled by the total number of stories. A higher building tends to produce more significant component deterioration to enlarge the maximum story drift angle at lower stories. The width-to-thickness ratio of beam and column components overtly affects the collapse capacity when the plastic deformation extensively develops. The ratio of residual to maximum story drift angle is significantly sensitive to the collapse capacity of various building models. A thin-walled concrete filled steel tubular (CFST) column is proposed as one efficient alternative to enhance the overall stiffness and deformation capacity of the high-rise SMRFs with fragile collapse performance. With the equivalent flexural stiffness, CFST-MRF buildings with thin-walled members demonstrate higher capacity to avoid collapse, and the greater collapse margin indicates that CFST-MRFs are a reasonable system for high-rises in seismic prone regions.
Tree boosting is a highly effective and widely used machine learning method. In this paper, we describe a scalable end-to-end tree boosting system called XGBoost, which is used widely by data scientists to achieve state-of-the-art results on many machine learning challenges. We propose a novel sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning. More importantly, we provide insights on cache access patterns, data compression and sharding to build a scalable tree boosting system. By combining these insights, XGBoost scales beyond billions of examples using far fewer resources than existing systems.