Structural Health Monitoring

During practical applications, the time-varying service conditions usually lead to difficulties in properly interpreting structural health monitoring signals. The guided wave–hidden Markov model–based damage evaluation method is a promising approach to address the uncertainties caused by the time-varying service condition. However, researches that have been performed to date are not comprehensive. Most of these research studies did not introduce serious time-varying factors, such as those that exist in reality, and hidden Markov model was applied directly without deep consideration of the performance improvement of hidden Markov model itself. In this article, the training stability problem when constructing the guided wave–hidden Markov model initialized by usually adopted k-means clustering method and its influence to damage evaluation were researched first by applying it to fatigue crack propagation evaluation of an attachment lug. After illustrating the problem of stability induced by k-means clustering, a novel uniform initialization Gaussian mixture model–based guided wave–hidden Markov model was proposed that provides steady and reliable construction of the guided wave–hidden Markov model. The advantage of the proposed method is demonstrated by lug fatigue crack propagation evaluation experiments.

This paper presents a temperature compensation method for Lamb wave structural health monitoring. The proposed approach considers a representation of the piezo-sensor signal through its Hilbert transform that allows one to extract the amplitude factor and the phase shift in signals caused by temperature changes. An ordinary least square (OLS) algorithm is used to estimate these unknown parameters. After estimating these parameters at each temperature in the operating range, linear functional relationships between the temperature and the estimated parameters are derived using the least squares method. A temperature compensation model is developed based on this linear relationship that allows one to reconstruct sensor signals at any arbitrary temperature. The proposed approach is validated numerically and experimentally for an anisotropic composite plate at different temperatures ranging from [Formula: see text] to [Formula: see text]. A close match is found between the measured signals and the reconstructed ones. This approach is interesting as it needs only a limited set of piezo-sensor signals at different temperatures for model training and temperature compensation at any arbitrary temperature. Damage localization results after temperature compensation demonstrate its robustness and effectiveness.

This paper presents a novel approach to detect structural damage based on combining independent component analysis (ICA) extraction of time domain data and artificial neural networks (ANN). The advantage of using time history measurements is that the original vibration information is used directly. However, the volume of data, measurement noise and the lack of reliable feature extraction tools are the major obstacles. To circumvent them, the independent component analysis technique is applied to represent the measured data with a linear combination of dominant statistical independent components and the mixing matrix [ A]. Such a representation captures the essential structure of the measured vibration data. The vibration features represented by the mixing matrix provide the relationship between the measured vibration response and the independent components and are then employed to build the simplified neural network model for damage detection. Two examples are included to demonstrate the effectiveness of the method. First, a truss structure with simulated displacement data was used, and the results show that healthy and damage states located in the nine elements may be classified. Second, a bookshelf structure together with measured time history data from 24 piezoelectric single axis accelerometers was used to demonstrate the approach on a physical structure. The results show the successful detection of the undamaged and damaged states with very good accuracy and repeatability.

Wayside detection monitors critical parameters relating to the condition of in-service railway vehicles. Economic decisions about the maintenance of vehicles can be made, and servicing can occur when a particular vehicle is likely to cause even small amounts of damage to the track, to itself, or when the cost of damage is significant, such as in catastrophic failure.

Vehicles with poorly performing axle bearings, out-of-round (skidded or spalled) wheels, vehicles which exhibit transient lateral motion (‘hunting’), and vehicles with poorly performing brakes are all likely to fall into the category of requiring maintenance, in order to save the track and the vehicle owner's money.

In the present paper, the parameters that define vehicle condition and their measurable effects are stated. There are frequently a number of wayside detection methods of inspecting a vehicle for the same vehicle condition and each of these is described in detail.

This investigation reveals the need for further research to enable rollingstock owners to make better decisions about the cost of operating their vehicles, based on the output from wayside detectors and the observed trends in wheel impact.

Deep learning algorithm can effectively obtain damage information using labeled samples, and has become a promising feature extraction tool for ultrasonic guided wave detection. But it is difficult to apply the monitoring expertise of structure A to structure B in most cases due to the differences in the dispersion and receiving modes of different waveguides. For multi-structure monitoring at the system level, how to transfer a trained structural health monitoring model to another different structure remains a major challenge. In this article, a cross-structure ultrasonic guided wave structural health monitoring method based on distribution adaptation deep transfer learning is proposed to solve the feature generalization problem in different monitoring structures. First, the joint distribution adaptation method is employed to adapt both the marginal distribution and conditional distribution of the guided wave signals from different structures. Second, convolutional long short-term memory network is constructed to learn the mapping relationship from adapted training samples in source domain. Batch normalization layer is implemented to balance the input tensors of each sample to the same distribution. Finally, the multi-sensor damage indexes are utilized to visually present the damage by probability imaging. The experimental results show that proposed method can utilize the single-sensor monitoring data in one structure to implement the multi-sensor damage monitoring in another structure and achieve the damage imaging visualization. The imaging performance is significantly superior to the existing principal component analysis, transfer component analysis, and other state-of-art comparison methods.

The centrifugal pump has a long history of frequent failure of anti-friction bearing since its commissioning in 1985. Vibration based conventional condition monitoring has regularly been used to identify the progressive nature of the bearing failure, but has failed to identify the root cause. Modal tests have been conducted on the pump assembly to understand the dynamics of the complete assembly. A typical case of the resonance of the bearing pedestals with 2X component (two times the pump RPM) of the response during pump operation mainly due to nonlinear interaction between the pump foundation and the concrete floor has been identified as the main source of the bearing failure. The results, observations, and the diagnosis to identify the root cause are discussed here.

The bearing vibration signal possesses nonlinear and non-stationary characteristics; hence; it is difficult to diagnosis the faults in the bearing under different working conditions. In this article, a new scheme has been proposed based on complete ensemble empirical mode decomposition with adaptive noise and corrected conditional entropy to recognize the different class of faults in bearing. The mode with minimum corrected conditional entropy is treated as a prominent mode from which sensitive features are extracted. A filter-based feature selection scheme is used for the same and for ranking the features based on variance to reduce the redundancy of data set. This data set is made input to support vector machine. The performance of the support vector machine classifier is improved by optimizing its parameters to obtain maximum classification accuracy. To address this issue, an evolutionary algorithm (diversity-driven multi-parent evolutionary algorithm) is used. With optimized support vector machine parameters, the support vector machine is trained to build a classification model with 10-fold cross-validation. After training, the built model is tested against test data set for fitness evaluation. The support vector machine classifier gives 100% accuracy at regularization and kernel parameter’s value of 1.3343 and of 782.6329, respectively, with 27.93 s of training time for a single iteration.

Fused deposition modeling is a popular technique for three-dimensional prototyping since it is cost-effective, convenient to operate, and environment-friendly. However, the low quality of its printed products jeopardizes its future development. Distortion, also known as warping deformation, which is caused by many factors such as inappropriate process parameters and process drifts, is one of the most common defects in the fused deposition modeling process. Rapid detection of such part distortion during the printing process is beneficial for improving the production efficiency and saving materials. In this article, a real-time part distortion monitoring method based on acoustic emission is presented. Our work is to identify distortion defects and understand the condition of the distortion area through sensing and digital signal processing techniques. In our experiments, both the acoustic emission hits and original signals were acquired during the fused deposition modeling process. Then, the acoustic emission hits were analyzed. Ensemble empirical mode decomposition was utilized to eliminate noise and extract features from the original acoustic emission signal to further analyze the acoustic emission signal in the case of part distortion. Furthermore, the root mean square of the reconstructed signals was calculated, and the prediction results are strongly correlated with the ground truth printing states. This work provides a promising method for the quality diagnosis of printing parts.

In this research, we attempt to establish a reliable structural health monitoring technique for composite materials by combining phase-shifted fiber-optic Bragg grating sensing with the laser ultrasonic visualization technology. In the first part of this article, a novel cross-adhesion configuration is designed to resolve the directionality problem of the phase-shifted fiber-optic Bragg grating ultrasonic sensing. In the adhesion configuration, Lamb waves are guided by an orthogonally bonded optical fiber from the adhesion point to the phase-shifted fiber-optic Bragg grating sensor. The analysis of the ultrasonic measurement results reveals that the proposed adhesion method enables us to use one sensor to receive Lamb waves in all in-plane directions with similar magnitude because two wave components propagating along with the two orthogonal directions are guided to the phase-shifted fiber-optic Bragg grating sensor and exhibit a linear superposition in the sensor. This simplified configuration gives our method an advantage over the existing approaches, such as the rosette configuration in which three or more phase-shifted fiber-optic Bragg grating sensors are required to relieve the sensing directionality. The phase-shifted fiber-optic Bragg grating ultrasonic sensor with the proposed adhesion configuration is then applied to visualize the propagation of ultrasonic waves in aluminum plates and carbon fiber–reinforced plastic laminates. Those verification experiments also show us that the new adhesion configuration is effective at protecting the phase-shifted fiber-optic Bragg grating ultrasonic measurement from the sensing directionality. Meanwhile, the broad bandwidth of the phase-shifted fiber-optic Bragg grating sensor enables us to visualize the propagation behavior of various Lamb wave modes over a broad frequency range. Finally, we also validate that the ultrasonic visualization technique merged with the phase-shifted fiber-optic Bragg grating ultrasonic sensing can be used to identify the hidden damage in the carbon fiber–reinforced plastic composite.