Structural Response Signals Research Articles

A Novel Classification Method for Flutter Signals Based on the CNN and STFT

Necessary model calculation simplifications, uncertainty in actual wind tunnel test, and data acquisition system error altogether lead to error between a set of actual experimental results and a set of theoretical design results; wind tunnel test flutter data can be utilized to feedback this error. In this study, a signal processing method was established to use the structural response signals from an aeroelastic model to classify flutter signals via deep learning algorithm. This novel flutter signal processing and classification method works by combining a convolutional neural network (CNN) with time-frequency analysis. Flutter characteristics are revealed in both time and frequency domains, which are harmonic or divergent in the time series; the flutter model energy is singular and significantly increases in the frequency view, so the features of the time-frequency diagram can be extracted from the dataset-trained CNN model. As the foundation of the subsequent deep learning algorithm, the datasets are placed into a collection of time-frequency diagrams calculated by short-time Fourier transform (STFT) and labeled with two artificial states, flutter or no flutter, depending on the source of the signal measured from a wind tunnel test on the aeroelastic model. After preprocessing, a cross-validation schedule is implemented to update (and optimize) CNN parameters though the trained dataset. The trained models were compared against test datasets to validate their reliability and robustness. Our results indicate that the accuracy rate of test datasets reaches 90%. The trained models can effectively and automatically distinguish whether or not there is flutter in the measured signals.

Research on modal analysis method of CNC machine tool based on operational impact excitation

With the rapid development of modern industry and industrial upgrading, machine tool performances such as high machining speed, high machining accuracy, and high reliability have gradually become the development trend of high-end CNC machine tools. The vibration characteristics of the machine structure become an important factor affecting the machining quality and processing efficiency of the machine tool. Since the movable parts of the CNC machine tool will impact the machine structure itself during the acceleration and deceleration process, by identifying the structural vibration response signal, the modal parameters of the machine tool in the running state can be obtained, called active excitation mode analysis. However, this method still has certain defects in theory and experiment. For example, the excitation principle is not theoretically explained, and the excitation energy and the frequency band of the method are limited. To solve these problems, this paper conducts more comprehensive research on the modal analysis method of CNC machine tools based on operational impact excitation. The main research contents are as follows: considering the rigid body motion of the table and the elastic collision of the screw nut pair, the impact excitation is modeled and analyzed. We have introduced a response-based modal identification algorithm. The influence factors of impact excitation energy and frequency band were studied from both theoretical and experimental aspects. Design three excitation sequences for workbench excitation and spindle excitation, respectively, analyze the vibration response and modal identification of the machine tool structure under different excitation sequences. It is verified that the method does not depend on the specific excitation sequence, as long as it satisfies the random characteristics. Aiming at the problem of insufficient excitation energy of the machine tool structure for single-component impact excitation, a multi-component joint impact excitation method based on the principle of multi-point excitation technology is proposed. The design experiment is carried out to verify the impact excitation method, which shows that the joint excitation has better excitation effect than single-axis excitation.

Aging State Detection of Viscoelastic Sandwich Structure Based on ELMD and Sensitive IA Spectrum Entropy

Viscoelastic sandwich structure is playing an important role in mechanical equipment, but therein viscoelastic material inevitably suffers from aging which affects structural service performance and the whole performance of equipment. Therefore, the aging state detection of viscoelastic sandwich structure based on vibration response signal is essential for monitoring the health state of structure and guaranteeing the operation safety of equipment. However, the weakness of structural vibration response variation caused by material aging make this task challenging. In this paper, a novel method based on ensemble local mean decomposition (ELMD) and sensitive IA spectrum entropy is proposed for this task. As an adaptive nonlinear and non-stationary signal processing method, ELMD is introduced to decompose the structural vibration response signal, and a series of instantaneous amplitudes (IAs) are obtained. Then, the spectrum entropies of these IAs are developed to quantitatively assess the aging state of viscoelastic sandwich structure. However, the IA spectrum entropies have different sensitivities to the aging state. Therefore, the most sensitive IA spectrum entropy is selected with a distance evaluation technique to detect the aging state of viscoelastic sandwich structure. In order to demonstrate the effectiveness of the proposed method, the experimental device of a viscoelastic sandwich structure is designed, and different structural aging states are created through the accelerated aging of viscoelastic material. The results show the outstanding performance of the proposed method.

Damage degree prediction method of CFRP structure based on fiber Bragg grating and epsilon-support vector regression

The assessment of structural damage is of great significance for ensuring the service safety of carbon fiber reinforced plastics (CFRP) structures. In this paper, the damage degree prediction method of CFRP structure based on fiber Bragg grating and epsilon-support vector regression was studied. The structural dynamic response signals were detected by fiber Bragg grating sensors. Then, the Fourier transform was used to extract the dynamic characteristics of the structure as the damage feature, and the damage feature dimensionality was reduced by using the RReliefF algorithm. On this basis, the damage degree prediction model of CFRP structure based on epsilon-support vector regression was established. Finally, the method proposed in this paper was experimentally verified. The results showed that the epsilon-support vector regression model can accurately predict the damage degree of unknown samples, and the absolute relative error of 27 experiments was less than 10% for 30 testing experiments. This paper provided a feasible method for predicting the damage degree of CFRP structures.

Full Dynamic Compound Inverse Method for seismic output-only element-level and input identification: Unitary formulation and extensive validation

Abstract The present work develops a unitary formulation of a novel deterministic output-only system identification and input estimation technique named Full Dynamic Compound Inverse Method (FDCIM) and outlines a systematic investigation on its consistent adoption within seismic engineering, by considering specific aspects of Signal Processing and producing a complete statistical evaluation. The achieved comprehensive FDCIM technique successfully estimates, all together: modal properties, element-level structural characteristics and earthquake input time history. Here, previous detached seminal contributions are first rejoined, in order to deal with various forms of structural viscous damping. Then, a wide database of ten selected real seismic excitations is taken into account to generate a set of synthetic earthquake-induced structural response signals, from a benchmark low-rise three-storey shear-type frame and from a realistic mid-rise ten-storey building. Such synthetic signals, possibly affected by noise of a controlled amount, are successfully processed via FDCIM and the achieved results demonstrate, as a first necessary validation condition, the consistency of the present identification algorithm toward potential Structural Health Monitoring purposes within the earthquake engineering range. Finally, a closing validation is proposed with the use of real earthquake-induced response signals, from an instrumented case study out of the CESMD database, with rather promising identification outcomes.

Research on FBG-Based CFRP Structural Damage Identification Using BP Neural Network

A damage identification system of carbon fiber reinforced plastics (CFRP) structures is investigated using fiber Bragg grating (FBG) sensors and back propagation (BP) neural network. FBG sensors are applied to construct the sensing network to detect the structural dynamic response signals generated by active actuation. The damage identification model is built based on the BP neural network. The dynamic signal characteristics extracted by the Fourier transform are the inputs, and the damage states are the outputs of the model. Besides, damages are simulated by placing lumped masses with different weights instead of inducing real damages, which is confirmed to be feasible by finite element analysis (FEA). At last, the damage identification system is verified on a CFRP plate with 300 mm × 300 mm experimental area, with the accurate identification of varied damage states. The system provides a practical way for CFRP structural damage identification.

STATISTICAL PATTERN RECOGNITION FOR STRUCTURAL HEALTH MONITORING USING ESN FEATURE EXTRACTION METHOD

For structural dynamic response signal, as the traditional featureextraction methods based on statistical pattern recognition arelinear and insensitive, this paper proposes a new signal featureextraction method based on the echo state network (ESN).

Assessment of Frequency versus Time Domain enhanced technique for response-only modal dynamic identification under seismic excitation

In the present study, response-only system identification is explored towards estimating modal dynamic properties of buildings under earthquake excitation. Seismic structural response signals are adopted with two different implemented Operational Modal Analysis (OMA) techniques, namely a refined Frequency Domain Decomposition algorithm and an improved data-driven Stochastic Subspace Identification (SSI-DATA) procedure, working in the Frequency Domain and in the Time Domain, respectively. These algorithms are specifically conceived to operate with seismic structural responses and at simultaneous heavy damping (in terms of identification challenge), towards achieving consistent estimations of natural frequencies, mode shapes and modal damping ratios. Classical OMA assumptions shall not contemplate short-duration, non-stationary earthquake-induced response signals. Nevertheless, the present enhanced output-only algorithms allow for estimating all strong ground motion modal parameters. First, a linear three-storey frame structure under ten selected earthquake base-excitations is considered, in order to assess the two OMA techniques at synthetic seismic response input. Comprehensive results are presented, by comparing the two developed methods, between achieved modal estimates and sought target values. Second, an existing instrumented building (CESMD database) is analyzed, by adopting real seismic response signals, reaching again effective modal estimates and corroborating the previous necessary condition analysis. According to this investigation, best up-to-date, re-interpreted, output-only techniques may effectively be used for potential Structural Health Monitoring purposes in the Earthquake Engineering range.

Earthquake‐induced structural response output‐only identification by two different Operational Modal Analysis techniques

SummaryOutput‐only system identification is developed here towards assessing current modal dynamic properties of buildings under seismic excitation. Earthquake‐induced structural response signals are adopted as input channels for two different Operational Modal Analysis (OMA) techniques, namely, a refined Frequency Domain Decomposition (rFDD) algorithm and an improved Data‐Driven Stochastic Subspace Identification (SSI‐DATA) procedure. Despite that short‐duration, non‐stationary, earthquake‐induced structural response signals shall not fulfil traditional OMA assumptions, these implementations are specifically formulated to operate with seismic responses and simultaneous heavy damping (in terms of identification challenge), for a consistent estimation of natural frequencies, mode shapes, and modal damping ratios. A linear ten‐storey frame structure under a set of ten selected earthquake base‐excitation instances is numerically simulated, by comparing the results from the two identification methods. According to this study, best up‐to‐date, reinterpreted OMA techniques may effectively be used to characterize the current dynamic behaviour of buildings, thus allowing for potential Structural Health Monitoring approaches in the Earthquake Engineering range.

Operational modal analysis of three-dimensional structures by second-order blind identification and least square generalized inverse

Most complex engineering structures are three-dimensional in practice. The process of one-dimensional extending to three-dimensional is a challenge that must be conquered by Operational Modal Analysis (OMA) methods when these methods are applied to complex engineering applications supported by scientific researches. This study put forward a new three-dimensional structure OMA method based on Second-Order Blind Identification (SOBI) and general reversion of least square. Firstly, modal coordinates decomposition of one-dimensional structural vibration response signal with SOBI. Secondly, the reasons that modal parameters identified by SOBI including energy uncertainty, order uncertainty and modal missing are explained in theory. Thirdly, the SOBI algorithm is used to decompose the response signals of displacement of a direction whose vibration response is the largest, then the other two directions are calculated by using the least square generalized inverse algorithm, and the modal parameters of three-dimensional structures are identified by the matrix assembly method. Numerical simulation results in a cylindrical shell demonstrated that this novel method is practical and effective by applied to practice in OMA of three-dimensional structures, and robustness to Gauss measurement noise disturbances.

Seismic FDD modal identification and monitoring of building properties from real strong-motion structural response signals

In the present study, output-only modal dynamic identification and monitoring of building properties is attempted successfully by processing real earthquake-induced structural response signals. This is achieved through an enhanced version of a recently-developed refined Frequency Domain Decomposition (rFDD) approach, which in the earlier implementation was adopted to analyse synthetic seismic response signals only. Despite that short duration, nonstationary seismic response data and heavy structural damping shall not fulfil traditional Operational Modal Analysis assumptions, the present rFDD response-only algorithm allows for the effective estimation of strong-motion natural frequencies, mode shapes, and modal damping ratios, with real seismic response signals. The present rFDD enhancement derives from a preprocessing time-frequency analysis and from an integrated approach for Power Spectral Density matrix computation, which constitute crucial innovative issues for the treatment of real earthquake response data. A monitoring case study is analysed by taking the real strong-motion response records from a seven-storey reinforced concrete building in Van Nuys, California, from 1987 to the latest 2014 events (Center of Engineering Strong Motion Data database), as recorded before, during and after the 1994 Northridge earthquake, which severely damaged the building (then retrofitted). This paper proves the effectiveness of the proposed enhanced rFDD algorithm as a robust method for monitoring current structural modal properties under real earthquake excitations. This shall allow for identifying possible variations of structural features along experienced seismic histories, providing then a fundamental tool towards Earthquake Engineering and Structural Health Monitoring purposes.

Full Dynamic Compound Inverse Method: Extension to General and Rayleigh damping

The present paper takes from the original output-only identification approach named Full Dynamic Compound Inverse Method (FDCIM), recently published on this journal by the authors, and proposes an innovative, much enhanced version, in the description of more general forms of structural damping, including for classically adopted Rayleigh damping. This has led to an extended FDCIM formulation, which offers superior performance, on all the targeted identification parameters, namely: modal properties, Rayleigh damping coefficients, structural features at the element-level and input seismic excitation time history. Synthetic earthquake-induced structural response signals are adopted as input channels for the FDCIM approach, towards comparison and validation. The identification algorithm is run first on a benchmark 3-storey shear-type frame, and then on a realistic 10-storey frame, also by considering noise added to the response signals. Consistency of the identification results is demonstrated, with definite superiority of this latter FDCIM proposal.

Plate-like structure damage location identification based on Lamb wave baseline-free probability imaging method

Damage-scattering signal extraction using conventional ultrasonic guided wave–based damage detection techniques requires the measurement of baseline data under pristine condition. This study proposes a baseline-free ultrasonic guided wave damage localization and imaging method based on Lamb wave baseline-free probability imaging method. Although traditional Lamb wave probability imaging can monitor damage location in plate-like structures, the absolute time of arrival and magnitude of the signal are affected by several factors and are therefore difficult to obtain. This study also proposes a probability-based hyperbola diagnostic imaging method that is based on different times of arrival and has no magnitude information. A distributed active sensor network conforming to a pulse-echo configuration and time window functions is developed to separate damage-scattering signals from structural response signals. Continuous wavelet transform is used to calculate the time of flight of damage signal waves. The numerical simulation and experiments validate the effectiveness of the proposed method in identifying damage.

Mono-modal feature extraction for bonding quality detection of explosive clad structure with optimized dual-tree complex wavelet transform

Bonding quality detection of explosive clad structure is significant to prevent catastrophic accidents. Multi-modal features related to bonding quality are contained in structural vibration response signal. Different modal feature has different sensitivity to the bonding quality. Extracting the desired mono-modal feature from the vibration response signal is necessary. Due to the mode aliasing easily appeared in the process of extracting the desired mono-modal feature, there is no effective method for this task. Dual-tree complex wavelet with attractive properties such as shift invariance and reduced spectral aliasing may provide a better way to extract the mono-modal feature. However, the fixed basis functions independent of the analyzed signal may weak the advantage of the method and even reduce the accuracy of detection result. To overcome this shortcoming, a technique called optimized dual-tree complex wavelet transform (ODTCWT) is proposed in this paper. Based on the analyzed signal, the optimized dual-tree complex wavelet basis function is constructed by searching for the proper parameters of vanishing moment K and the order of filter L. The optimized dual-tree complex wavelet with improved wavelet filters can best matched the modal frequencies of the analyzed signal. The ODTCWT can extract the mono-modal feature from vibration response signal with lower mode aliasing. The feasibility and effectiveness of the method of constructing ODTCWT is illustrated by the simulated signal. The proposed ODTCWT is combined with time entropy to detecting bonding quality of explosive clad pipes. For comparison, un-optimized dual-tree complex wavelet transform (UODTCWT), second-generation wavelet transform (SGWT) and band-pass filter (BPF) are also used for this task to demonstrate the validity of ODTCWT.

Compressive sampling and data fusion-based structural damage monitoring in wireless sensor network

The Lamb wave phased array structural health monitoring method is effective in structural damage monitoring. In this method, the damage scattering signal can be obtained by comparing the damage structural response signal with health structural response signal, and it can be used for structural damage identification. But in the structural health monitoring based on wireless sensor networks, this method has some inevitable defects in data transmission. A large number of sampling data of damage response signal will cause huge wireless communication burden. To solve this problem, we proposed a phased array image method based on compressive sampling and data fusion for wireless structural damage monitoring. First, compressive sampling signal by compressive sampling method was collected. Then, data fusion for multi-sensor’s damage response signal was implemented in phased array. Finally, the Lamb wave phased array damage identification method based on compressive sampling and data fusion was proposed. Experimental results on carbon composite structure show that the proposed method can largely save network bandwidth and energy. This method can also realize the damage identification accurately on the aviation aluminum plate and keep the detection error within 0.82 mm.

Simplified fuzzy ARTMAP network-based method for assessment of structural damage applied to composite structures

This paper addresses a novel method for structural damage assessment based on the electromechanical impedance (EMI) technique applied to composite structures. This method is based on a high-frequency excitation range in order to overcome the difficulties caused when the low vibration modes are excited. A structure made up of unidirectional carbon fiber pre-impregnated with resin-epoxy is excited using a chirp signal which varies the frequency in a wide range through the PZT (Lead Zirconate Titanate) patches. From the structural response signal, the Euclidean Distance (ED) is computed from each structural condition considering the pristine condition as a reference (baseline). The ED is used as input to the Simplified Fuzzy ARTMAP Network (SFAN) which is responsible for identifying various different structural conditions. It is shown that the robustness provided by SFAN allows the evaluation of the progress of structural damage even under critical conditions. The structural damage was simulated by loosening bolts in a structure. Furthermore, the repaired condition was also considered by retightening bolts. The paper discusses the advantages and drawbacks of the approach in light of the experimental results.

A Full Dynamic Compound Inverse Method for output-only element-level system identification and input estimation from earthquake response signals

This paper proposes a new output-only element-level system identification and input estimation technique, towards the simultaneous identification of modal parameters, input excitation time history and structural features at the element-level by adopting earthquake-induced structural response signals. The method, named Full Dynamic Compound Inverse Method (FDCIM), releases strong assumptions of earlier element-level techniques, by working with a two-stage iterative algorithm. Jointly, a Statistical Average technique, a modification process and a parameter projection strategy are adopted at each stage to achieve stronger convergence for the identified estimates. The proposed method works in a deterministic way and is completely developed in State-Space form. Further, it does not require continuous- to discrete-time transformations and does not depend on initialization conditions. Synthetic earthquake-induced response signals from different shear-type buildings are generated to validate the implemented procedure, also with noise-corrupted cases. The achieved results provide a necessary condition to demonstrate the effectiveness of the proposed identification method.

Multi-Damage Identification System of CFRP by Using FBG Sensors and Multi-Classification RVM Method

Multi-damage identification of carbon fiber reinforced plastics (CFRP) structure is very important to ensure the structural safety. In this paper, the multi-damage identification system of CFRP using fiber Bragg grating (FBG) sensors and multi-classification relevance vector machine (RVM) method was proposed. First, the multi-damage identification method was researched by an active actuation method. In addition, the structural dynamic response signals were detected by the FBG sensors. Then, the multi-damage characteristics extraction was completed by Fourier transform, reliefF, and principal component analysis algorithms. With the multi-damage characteristics as the input and their corresponding multi-damage state as the output, the multi-classification RVM model was trained to identify the multi-damage state. Finally, the multi-damage identification system was established and verified on a CFRP plate with dimensions of 500 mm $\times \,\, 500$ mm $\times \,\, 2$ mm. The results showed that the proposed multi-damage identification method can accurately identify the CFRP structural multi-damage state. This paper provided a reliable method for the CFRP structural multi-damage identification.

Damage identification system of CFRP using fiber Bragg grating sensors

Carbon fiber reinforced plastics (CFRP) structure is very sensitive to impact damage, thus an early identification of structural damage is necessary to avoid major structural failures. In this paper, structural damage identification system by using fiber Bragg grating (FBG) sensors and damage identification algorithm for CFRP structure were investigated. Structural dynamic response signals, which contain the damage information, were generated by an active actuation method, and detected by FBG sensors. Damage characteristics were extracted by the Fourier transform and principal component analysis (PCA) methods. Then, based on the damage characteristic, one-class support vector machines (SVM) method was used to detect the damage, and multi-class C-support vector classification (C-SVC) methods were used to identify the damage area and evaluate the damage degree. At last, the damage identification system was established and verified on a CFRP plate. Within a 160mm∗160mm∗2mm experiment area, the identification accuracy was more than 90%. This paper provided a reliable method for the CFRP structural damage identification.

CFRP damage identification system based on FBG sensors and ELM method

The identification of the damage state of Carbon fiber-reinforced plastic (CFRP) structure is the necessary information for ensuring the safety of CFRP structure. In this paper, the structural damage identification system using fiber Bragg grating (FBG) sensors and the damage identification method were investigated. FBG sensors were used to detect the structural dynamic response signal, which was generated by an active actuation way. Fourier transform and principal component analysis (PCA) were used to extract the damage characteristic. After that, the structural damage identification model was constructed based on extreme learning machine (ELM), whose input is the damage characteristic and output is the damage state. Finally, the damage identification system was established and verified on a CFRP plate with 160 mm $$\,\times$$ 160 mm experiment area. The experimental results showed that the identification accuracy was more than 90 %. This paper provided a reliable method for CFRP structural damage identification.