Parallel implementation of the time domain spectral element method on GPU (Graphics Processing Unit) is presented. The proposed spectral element method implementation is based on sparse matrix storage of local shape function derivatives calculated at Gauss–Lobatto–Legendre points. The algorithm utilizes two basic operations: multiplication of sparse matrix by vector and element-by-element vectors multiplication. Parallel processing is performed on the degree of freedom level. The assembly of resultant force is done by the aid of a mesh coloring algorithm. The implementation enables considerable computation speedup as well as a simulation of complex structural health monitoring systems based on anomalies of propagating Lamb waves. Hence, the complexity of various models can be tested and compared in order to be as close to reality as possible by using modern computers. A comparative example of a composite laminate modeling by using homogenization of material properties in one layer of 3D brick spectral elements with composite in which each ply is simulated by separate layer of 3D brick spectral elements is described. Consequences of application of each technique are explained. Further analysis is performed for composite laminate with delamination. In each case piezoelectric transducer as well as glue layer between actuator and host structure is modeled.
Various non-destructive testing (NDT) methods have been developed to extract information about state of a structure. Two of them: vibration-based and guided wave-based techniques are one of the most commonly used and well developed. Both approaches can be implemented using Scanning Laser Doppler Vibrometer measurements and excitation by means of piezoelectric transducer. In this paper authors present a combined approached for NDT using successive and simultaneous measurement of both mode shapes and guided waves. Vibration-based damage detection is focused on detection of mode shape singularity, created by material discontinuity. This method utilizes wavelet transform and Teager energy operator for damage indication. Guided wave-based damage detection uses propagating elastic wave energy variation on the specimen surface as well as any changes in wave propagation pattern due to its interaction with material discontinuity as a tool for structural health assessment. Combining this two different techniques can give higher accuracy in defect detection. At the same time any additional specimen preparation are necessary, any set-up changes are required and the all the data can be registered in the same amount of time (simultaneous excitation). To confirm proposed technique a honeycomb core sandwich aluminum plate with debonding is tested. A results obtained with both techniques and combined approach are presented.
Proc. SPIE. 9805, Health Monitoring of Structural and Biological Systems 2016
KEYWORDS: Ferroelectric materials, Fiber Bragg gratings, Sensors, Fiber optics sensors, Data acquisition, Data processing, Laser Doppler velocimetry, Structural health monitoring, Modal analysis, Damage detection, Microsoft Foundation Class Library, Beam shaping, Data fusion
This work addresses the severe lack of literature in the area of modal analysis for multi-metric sensing. The paper aims at providing a step by step tutorial for performance of modal analysis using Fiber Bragg Grating (FBG) strain sensors and Laser Doppler Vibrometer (LDV) for displacement measurements. The paper discusses in detail the different parameters which affect the accuracy of the experimental results. It highlights the often implied, and un-mentioned problems, that researchers face while performing experiments. The paper tries to bridge the gap between the theoretical idea of the experiment and its actual execution by discussing each aspect including the choice of specimen, boundary conditions, sensors, sensor position, excitation mechanism and its location as well as the post processing of the data. The paper may be viewed as a checklist for performing modal analysis in order to ensure high quality measurements by avoiding the systematic errors to creep in.
The aim of this paper is to present aspects of Lamb wave propagation in randomly oriented short fiber reinforce composites with delamination. Prediction of elastic constants is based on mechanics of composites, rule of mixture and total mass balance tailored to the spectral element mesh composed of 3D brick elements. Piezoelectric excitation as well as glue layer are taken into account. Complex full wave field includes multiple reflections at short fibers. This wave pattern is also obtained by the use of laser vibrometry confirming good quality of the model. Further studies are related to symmetrical and non-symmetrical delamination in respect to the thickness of the composite plate. Square delamination of the side length 10 mm is investigated. It has been found that reflections from delamination are mostly superimposed with reflections coming from short fibers. Hence, delamination detection by direct analysis of wave propagation pattern on the surface of the plate is ineffective. However, adaptive wavenumber filtering method overcome these difficulties and enables not only to detect the delamination but also is helpful for delamination size estimation. Moreover, the method is more effective if the full wavefield measurements are acquired on the surface of the plate which is closer to the delamination.
Many studies have been published in recent years on Lamb wave propagation in isotropic and (multi-layered) anisotropic structures. In this paper, adiabatic wave propagation phenomenon in a tapered composite panel made out of glass fiber reinforced polymers (GFRP) will be considered. Such structural elements are often used e.g. in wind turbine blades and aerospace structures. Here, the wave velocity of each wave mode does not only change with frequency and the direction of wave propagation. It further changes locally due to the varying cross-section of the GFRP panel.
Elastic waves were excited using a piezoelectric transducer. Full wave-field measurements using scanning Laser Doppler vibrometry have been performed. This approach allows the detailed analysis of elastic wave propagation in composite specimen with linearly changing thickness. It will be demonstrated here experimentally, that the wave velocity changes significantly due to the tapered geometry of the structure. Hence, this work motivates the theoretical and experimental analysis of adiabatic mode propagation for the purpose of Non-Destructive Testing and Structural Health Monitoring.
In recent years electromagnetic Terahertz (THz) radiation or T-ray has been increasingly used for nondestructive evaluation of various materials such as polymer composites and porous foam tiles in which ultrasonic waves cannot penetrate but T-ray can. Most of these investigations have been limited to mechanical damage detection like inclusions, cracks, delaminations etc. So far only a few investigations have been reported on heat induced damage detection. Unlike mechanical damage the heat induced damage does not have a clear interface between the damaged part and the surrounding intact material from which electromagnetic waves can be reflected back. Difficulties associated with the heat induced damage detection in composite materials using T-ray are discussed in detail in this paper. T-ray measurements are compared for different levels of heat exposure of composite specimens.
Curvature mode shape is an effective feature for damage detection in beams. However, it is susceptible to measurement
noise, easily impairing its advantage of sensitivity to damage. To deal with this deficiency, this study formulates an
improved curvature mode shape for multiple damage detection in beams based on integrating a wavelet transform (WT)
and a Teager energy operator (TEO). The improved curvature mode shape, termed the WT - TEO curvature mode shape,
has inherent capabilities of immunity to noise and sensitivity to damage. The proposed method is experimentally
validated by identifying multiple cracks in cantilever steel beams with the mode shapes acquired using a scanning laser
vibrometer. The results demonstrate that the improved curvature mode shape can identify multiple damage accurately
and reliably, and it is fairly robust to measurement noise.
The aim of this paper is to present a method for visualization thermally induced delamination in composite material based on guided wave propagation phenomenon. Tested specimen was submitted to short time period high temperature source, which generated thermal degradation. In particular, delamination in material occurred. This procedure simulates some real case scenarios damage like one cased by atmospheric discharge striking wind turbine blade. Proposed method utilizes processing of full wavefield data acquired by the Scanning Doppler Laser Vibrometer. Registered wavefield images are transformed to wavenumber domain where the wave propagation pattern is removed. In this way after transformation signal back to space domain it contains only information about changes in wave propagation and may be used for damage visualization. However, attenuation of waves cause that visualized anomalies has lower amplitudes with increased distance from the actuator. The proposed enhancement of signal processing algorithm enables quantification of the size of the damage. The enhancement is a technique for compensation of the wave attenuation so that the effects of structural damages have the same influence regardless of the location.
The main objective of this study is to present a novel method for damage detection in plate-type structures using twodimensional (2D) continuous wavelet transforms. For this purpose, the 2D Mexican wavelet is employed to remold the equation of motion for transverse vibration of a plate. The remolded vibration equation of a plate can serve as a multiscale damage detection scheme that characterizes damage using an indicator of multiscale pseudo-load. Effects of multiscale pseudo-load can pinpoint the location of the damage as well as revealing its configuration; moreover, the strong solid mechanics foundation of the method results in the identified damage with an explicit physical implication. The performance of the proposed technique is validated through an experimental program of using a scanning laser vibrometer (SLV) to measure the transverse deflection of an aluminum plate bearing a cross-like notch and an added small mass. The results confirm the robustness and superior capability of the proposed method in detecting damage in plate-type structures.
The aim of this paper is to present methods for enhancing damage visualization in structures based on wave propagation
phenomenon. The method utilizes filtering and processing of full wavefield acquired by the laser vibrometer. Laser
vibrometer allows to register full wavefield in elements of a structure instead of single point measurements acquired by
e.g. piezoelectric sensor. In this way new possibilities for Nondestructive Evaluation arise enabling visualization of
elastic waves interacting with various types of damages. Measurements obtained with a scanning laser vibrometer can be
combined with effective signal and imaging processing algorithms to support damage identification. In this paper new
method for wave filtering of propagating waves is tested on both numerical results and experimental data obtained from
laser vibrometry measurements of composite plates. Processing of signals registered at a rectangular grid of
measurement points covering inspected area of the plate involve 2D DFFT (Discrete Fast Fourier Transform),
wavenumber filtering and inverse DFFT. As a result new damage index is proposed and compared with other methods
like RMS and frequency-wavenumber filtering.