Kerscher, Pedro Jose PachecoSchmith, JeanMartins, Eduardo Augustode Figueiredo, Rodrigo Marques...
8页
查看更多>>摘要：The spark test method is a simple and low-cost method in which an operator with special skills observes the sparks emitted by a grinding wheel in contact with the steel in order to identify the material of the sample. However, the operator might classify erroneously two different steel materials with close spark characteristics. Therefore, we propose a method that extracts features from images captured from the spark test method and uses these features as input on machine learning models. The regression models predicted the carbon content of steel with 8% error while the classifiers had 82% of accuracy. The classifiers models had good results with few confusion points and regression models had low error. Regarding the confusion points, the regression algorithms could solve the misclassification by predicting the carbon content of the sample and increasing accuracy. The proposed method is suitable for real-time and shop floor applications.
查看更多>>摘要：In this paper, a modified Levenberg-Marquardt algorithm (MLMA) is proposed to localize the 'point of burst' of an explosive sound source over the range of (0.5-2500) m. The objective function for minimization is formulated through the time difference of arrival based multilateration approach. The developed method uses four exclusive steps to satisfy global convergence along with fast computational speed. The performance of the proposed method is validated through both indoor-outdoor experiments and simulation studies and compared to other well-known methods. The experimental results show that the non-iterative approaches perform satisfactorily only if the ratio of microphone spacing to source range r is greater than 0.30. However, the iterative approaches outperform non-iterative approaches for any r. It is also observed that the MLMA converges globally at least five times faster than other algorithms. The numerical simulation results also demonstrate that the MLMA provides optimal solutions at lower and higher noise thresholds.
查看更多>>摘要：The falling-film flow characteristics significantly influence the heat and mass transfer process in the horizontal tube absorber. This study experimentally investigates the effect of ultrasonic field on the flow mode transition and thickness distribution of the liquid film on the horizontal tube. The water-calcium chloride (CaCl2/water) was selected as the working fluid. The effects of the factors, such as the tube diameter, tube spacing, solution concentration, ultrasonic power, sound source distance and ultrasonic incidence angle, on the hydrodynamics of the liquid flow were analyzed. The results show that the mode transition Reynolds number increases with the increasing ultrasonic power, but reduces comparing to that without ultrasound. Moreover, in the column mode, the fluctuation range of the film thickness decreases with the increase of the ultrasound power; and the ultrasonic field help increase the average film thickness outside the tube and enhance the uniformity of liquid film.
查看更多>>摘要：A new experimental procedure for damage detection based on active acoustic emission was developed which is sensitive to damage depth. Two new damage indexes were proposed by analyzing the amplitudes of lowfrequency signal direct wave and high-frequency signal maximum component. The relations between damage depths of a steel plate and damage indexes were studied through numerical simulation considering different signal-to-noise ratios (SNR). It was found that the values of the damage indexes increased with the depth of damage, and the stability of the indexes was further analyzed by the coefficient of variation (Cov). Damage detection experiments for steel plates with different damage depths were carried out, and the results showed good agreement with the numerical simulation which further verifies the feasibility of the proposed method. The present method can provide basis for the engineering application of damage detection with different depths.
查看更多>>摘要：This study aims at investigating the early-age behaviour of textile-reinforced cementitious matrix composites. Six different configurations with two types of reinforcements (glass and carbon meshes) and three different reinforcement layers were tested. Experimental measurements were conducted using various measurement techniques: optical fibre sensors, thermocouples, thermal cameras, and Fourier transform infrared spectroscopy analysis. The different phases of expansion and shrinkage were monitored during the first 12 h and after 15 days from casting. The effects of air contact, textile, and mesh size of reinforcement grids were evaluated and discussed. Then, the effect of shrinkage on the mechanical tensile behaviour of these composites was identified and analysed. Moreover, a chemical analysis was performed to determine the influences of the constituents, the cement hydration process and related heat release on the early-age behaviour of these cementitious matrix composites.
查看更多>>摘要：Condition monitoring is a major part of predictive maintenance which monitors a particular condition in machinery to identify changes that could indicate a developing fault. It allows maintenance to be scheduled and preventive actions to be taken to reduce the failures. This study presents a new feature extraction method that is used to detect the faults of worm gears (WG) during the condition monitoring process under various operating conditions. In this study, an experimental setup that can operate under different operating conditions has been developed to obtain vibration and acoustic data. The feature extraction technique Common Spatial Pattern (CSP) has been used for the first time to detect the faults (wear, pitting and tooth breakage) of machinery from vibration and acoustic data. Fault detection and classification were performed with Artificial Neural Network (ANN), Support Vector Machine (SVM) and K-Nearest Neighbor (k-NN) methods based on CSP features obtained using vibration and acoustic signals. According to the classification performance results, ANN method has produced considerable high accuracies for two class and multiclass classification when compared with the Support Vector Machine (SVM), K-Nearest Neighbor (k-NN). Moreover, the ANN classification results have also been compared with the Convolutional Neural Networks (CNNs) results in the literature. Finally, the performance of CSP features was validated with the commonly used time and frequency domain features. The contribution of this work includes the first time usage of CSP features for fault detection which were extracted from vibration and acoustic data of an experimental WG set. Moreover, various fault types of WGs under changing loading and speed have been examined for the first time. The results show that ANN with CSP features could achieve excellent performances in condition monitoring of WGs under variable operating conditions.
查看更多>>摘要：The polyurethane-encapsulated lateral line-inspired MEMS vector hydrophone has problems such as lowfrequency sensitivity (10-80 Hz) loss and narrow working bandwidth. In order to eliminate these two problems, a new encapsulation with a stainless steel mesh cap structure was proposed. This paper first verified the sensitivity loss and narrow working bandwidth of the polyurethane encapsulation through experiments and simulations, analyzed the impact of the perforation rate, thickness, and aperture of the stainless steel mesh cap structure on the vector hydrophone through simulation, and finally made 8 types of stainless steel mesh cap with different apertures was tested experimentally. Finally, the experimental results show that the sensitivity of the stainless steel mesh encapsulation in the low frequency range (10-80 Hz) is improved by an average of 12 dB than that of the polyurethane encapsulation, and the working bandwidth has been increased from 10-210 Hz to 10-667 Hz.
Yang, JaehyunMostaghimi, HamidHugo, RonPark, Simon S....
17页
查看更多>>摘要：Pipeline monitoring provides operators with invaluable information regarding the potential risks that may pose threats to the integrity of the entire line. Pipeline leakage results in serious environmental and financial costs that can be avoided through leak detection systems. This study introduces a comprehensive leak monitoring system that allows leak detection, localization, and volume rate estimation in liquid pipelines installed above ground, simultaneously. To minimize the leak interpretation errors an artificial intelligence (AI)-based leak detection algorithm is developed. Pressure sensors are utilized to capture real-time variations of fluid pressure to localize pipeline leakage through the application of the pressure gradient intersection method. Vibrations of the pipeline are also acquired in real-time through accelerometers, the signals of which are then used to estimate the leak forces through the inverse dynamics of the pipeline between the leak location and the location of the accelerometers. This is achieved by developing a leak-induced vibration (LIV) model that simulates the dynamics of the pipe through a finite element (FE) vibration model. The transfer function of the pipe assembly is then used to design a Kalman filter. The Kalman filter predicts the leak forces and is used to estimate the fluid release through correlation analysis of the leak forces and leak volume rate, experimentally. A lab-scale experimental setup is manufactured to verify the dynamic LIV model and to test the proposed methodology. The performance of the proposed methodology shows 97 %, 96 %, and 92 % of accuracy on average for leak detection, localization, and leak volume rate estimation, respectively.
查看更多>>摘要：Novel hybrid carbon materials such as activated carbon-carbon nanotubes (AC-CNTs) and carbon spheres/activated carbon-carbon nanotubes (CSs/AC-CNTs) based on activated carbon (AC) derived from the pits of Algerian date palm have been prepared and characterized. In addition, the carbon paste electrodes based on graphite carbon (GC) and cavity microelectrode (CME) modified with these hybrid materials were used to detect gallic acid at pH = 7 using square wave voltammetry method (SWV). The conductivity measurements revealed that CSs/AC-CNTs is more conducting than AC-CNTs. SWV measurements showed that the oxidation current was directly proportional to the concentrations of gallic acid (from 0 to 0.00536 M) with the lowest limit of detection (LOD), reaching 6.43 mu M and 3.64 mu M using GC/CSs/AC-CNTs electrode and CME/GC/CSs/AC-CNTs sensor, respectively. The reproducibility and the stability of the studied sensor were confirmed by the relative standard deviation of the oxidation current response of gallic acid (RSD (Reproducibility) = 1.44% and RSD (Stability) = 3.7%).
查看更多>>摘要：We develop a novel sensor placement method that maximizes monitoring performance while minimizing deployment cost. Our method integrates a reduced order model and multi-objective combinatorial optimization. We first decompose the spatio-temporal state field to be monitored by proper orthogonal decomposition (POD), and we use the Gaussian Process to model the uncertainty in each POD mode. Next, we develop a lazy greedy (LG)-SMALL ELEMENT OF-constraint optimization to derive the Pareto-optimal sensor configurations. We further design a branch and bound algorithm to calculate the global optimum and validate the correctness of select configurations on the LG-derived Pareto frontier. We evaluate and benchmark our algorithm in computational experiments based on the temperature dataset of the Berkeley Intel Lab. The computational results demonstrate that our algorithm places sensors at locations of large magnitude in the POD modes, and that our method achieves better state estimation accuracy and smaller reconstruction errors compared with alternative methods.
NSTL
Elsevier