Filtern
Dokumenttyp
- Beitrag zu einem Tagungsband (8) (entfernen)
Sprache
- Englisch (8)
Schlagworte
- Guided waves (8) (entfernen)
Organisationseinheit der BAM
Understanding on how a machine learning model interprets data is a crucial step to verify its reliability and avoid overfitting. While the focus of the scientific community is nowadays orientated towards deep learning approaches, which are considered as black box approaches, this work presents a toolbox that is based on complementary methods of feature extraction and selection, where the classification decisions of the model are transparent and can be physically interpreted. On the example of guided wave benchmark data from the open guided waves platform, where delamination defects were simulated at multiple positions on a carbon fiber reinforced plastic plate under varying temperature conditions, the authors could identify suitable frequencies for further investigations and experiments. Furthermore, the authors presented a realistic validation scenario which ensures that the machine learning model learns global damage characteristics rather than position specific characteristics.
Understanding on how a machine learning model interprets data is a crucial step to verify its reliability and avoid overfitting. While the focus of the scientific community is nowadays orientated towards deep learning approaches, which are considered as black box approaches, this work presents a toolbox that is based on complementary methods of feature extraction and selection, where the classification decisions of the model are transparent and can be physically interpreted. On the example of guided wave benchmark data from the open guided waves platform, where delamination defects were simulated at multiple positions on a carbon fiber reinforced plastic plate under varying temperature conditions, the authors could identify suitable frequencies for further investigations and experiments. Furthermore, the authors presented a realistic validation scenario which ensures that the machine learning model learns global damage characteristics rather than position specific characteristics.
The use of composite materials has been steadily growing during the last decades, as well as the requirements on quality, mechanical properties and geometries of the parts. Some processes, like the 3D UV pultrusion process, manufacture parts that are long and thin, whilst having a varying curvature radius along the same part or even no curvature at all. Studying their mechanical properties along the main fiber direction, which is of foremost interest, is not an easy task nor efficient with most nondestructive methods. The use of air-coupled ultrasonics to evaluate the properties of composite materials has been widely proved by several authors, mainly using guided waves that provide information on the orthotropic properties of this kind of materials. Most of this work has focused on analyzing straight plate-like geometries, due to the simplicity to generate desired Lamb modes in the plate and analyze the behavior of guided waves inside the plate. In our contribution, the differences in the propagation of Lamb waves for straight and curved geometry glass fiber reinforced polymers (GFRP) have been analyzed. A GFRP test sample cured with UV light with one straight and one curved area has been evaluated. The responses of the generated Lamb wave modes for the straight and curved geometries have been compared, accounting for variations in the transducer characteristics, e.g. resonance behavior and focusing.
Guided waves are increasingly a subject of great interest in nondestructive testing. An example of research in this field is the development of a novel procedure for ultrasonic testing of wheelset-axles using guided waves, which requires to treat the wheelset-axle as a thick walled cylinder with varying thickness. In order to describe ultrasound propagation in a waveguide with non-constant thickness, a multimodal approach, which allows to avoid extensive mesh-based numerical calculations, seems to be promising. The method treats the modes of a corresponding waveguide with constant thickness as a base in which an arbitrary sound field can be expressed. Since the local sound field at any given position in the waveguide with varying thickness will be a combination of these base modes, the problem is reduced to solving the differential equation that governs the evolution of the coefficients in the mode spectrum along the waveguide. Once the description of the sound field along the waveguide is obtained, the time dependence is added by multiplication with a simple oscillating term. Simulations of pulse propagation through the waveguide can then be constructed by adding up a sufficient number of mono-frequent continuous wave solutions. As an early stage in developing a simulation tool for sound propagation in thick walled cylinders with varying thickness, the multimodal approach was implemented and tested for the simple case of plate geometries. In this work, an overview of the simulations carried out for plates with non-constant thickness is presented. The performance of the algorithm based on the multimodal approach and the obtained results are compared to those of mesh-based simulation tools.
Numerical simulation of ultrasonic guided waves using the scaled boundary finite element method
(2012)
The formulation of the Scaled Boundary Finite Element Method is applied for the computation of dispersion properties of ultrasonic guided waves. The cross-section of the waveguide is discretized in the Finite Element sense, while the direction of propagation is described analytically. A standard eigenvalue problem is derived to compute the wave numbers of propagating modes. This paper focuses on cylindrical waveguides, where only a straight line has to be discretized. Higher-order elements are utilized for the discretization. As examples, dispersion curves are computed for a homogeneous pipe and a layered cylinder.
Lamb waves travel in plates and hollow cylinders over large distances and propagate with multiple mode shapes. Therefore the waves can be used for integrity tests of large scale structures. Each propagating wave mode has a unique dispersive character and a frequency dependent vibration pattern. Therefore, by interacting with flaws each mode is assumed to generate an individual reflection pattern depending on the flaw type, flaw orientation and size. To extract this information a frequency as well as wave number dependent mode excitation is required ensuring a single mode excitation. The challenge, however, is to control the selective mode excitation electronically to allow a sequence of consecutive tests with different wave modes to collect the data. A method is proposed using phased array wedge transducers with fluid coupling for normal force excitation of Lamb wave modes. The excitation principle is investigated analytically on plates by modelling the wave generation in the wedge and the coupling between wedge and structure. The analytical results of the excitation principle are validated by corresponding experiments where the field patterns on the wedge surface and on the plate were scanned using a laser vibrometer. The directivity of the radiated field is investigated together with the direction dependent analysis of the mode content.
Guided waves travel in plates and hollow cylinders over large distances and propagate with multiple mode shapes. Therefore the waves can be used viably for integrity tests of large scale structures. The number of propagating modes increases with frequency. Due to their dispersive character the different modes are manageable only in a limited frequency range. Depending on the wave length and on the angle of impingement of the wave front to the coupling surface between transducer and structure, a trace wavelength is predefined and a selective excitation of single modes becomes feasible. By using phased array technique the excited wave mode can be selected by controlling the input signal of the transducer. Different modes are excitable with a sin-gle mechanical set-up. In a first step of the investigation, a calculation model is developed modelling the wave propagation and the selective excitation of guided wave modes depending on the control parameters. Dedicated experiments show the applicability of the method presented. The flaw detection of different sized cracks and of material thickness reductions is examined depending on the excitation wave mode.