Filtern
Dokumenttyp
- Beitrag zu einem Tagungsband (11)
- Zeitschriftenartikel (5)
- Vortrag (1)
- Forschungsdatensatz (1)
Schlagworte
- Lamb waves (5)
- Ultrasonic guided waves (4)
- Composites (3)
- Materialcharakterisierung (3)
- Adhesive Bonding (2)
- Dispersion (2)
- Geführte Ultraschallwellen (2)
- Impact damage (2)
- Inverse Probleme (2)
- Kissing Bonds (2)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (1)
Data-driven analysis for damage assessment has a large potential in structural health monitoring (SHM) systems, where sensors are permanently attached to the structure, enabling continuous and frequent measurements. In this contribution, we propose a machine learning (ML) approach for automated damage detection, based on an ML toolbox for industrial condition monitoring. The toolbox combines multiple complementary algorithms for feature extraction and selection and automatically chooses the best combination of methods for the dataset at hand. Here, this toolbox is applied to a guided wave-based SHM dataset for varying temperatures and damage locations, which is freely available on the Open Guided Waves platform. A classification rate of 96.2% is achieved, demonstrating reliable and automated damage detection. Moreover, the ability of the ML model to identify a damaged structure at untrained damage locations and temperatures is demonstrated.
The fourth dataset dedicated to the Open Guided Waves platform presented in this work aims at a carbon fiber composite plate with an additional omega stringer at constant temperature conditions. The dataset provides full ultrasonic guided wavefields.
A chirp signal in the frequency range 20-500 kHz and Hann windowed tone-burst signal with 5 cycles and carrier frequencies of 16.5 kHz, 50 kHz, 100 kHz, 200 kHz and 300kHz are used to excite the wave. The piezoceramic actuator used for this purpose is attached to the center of the stringer side surface of the core plate.
Three scenarios are provided with this setup: (1) wavefield measurements without damage, (2) wavefield measurements with a local stringer debond and (3) wavefield measurements with a large stringer debond. The defects were caused by impacts performed from the backside of the plate. As result, the stringer feet debonds locally which was verified with conventional ultrasound measurements.
The dataset can be used for benchmarking purposes of various signal processing methods for damage imaging.
The detailed description of the dataset is published in Data in Brief Journal.
The fourth dataset dedicated to the Open Guided Waves platform presented in this work aims at a carbon fiber composite plate with an additional omega stringer at constant temperature conditions. The dataset provides full ultrasonic guided wavefields. Two types of signals were used for guided wave excitation, namely chirp signal and tone-burst signal. The chirp signal had a frequency range of 20-500kHz. The tone-burst signals had a form of sine modulated by Hann window with 5 cycles and carrier frequencies 16.5kHz, 50kHz, 100kHz, 200kHz, 300kHz. The piezoceramic actuator used for this purpose was attached to the center of the stringer side surface of the core plate. Three scenarios are provided with this setup: (1) wavefield measurements without damage, (2) wavefield measurements with a local stringer debond and (3) wavefield measurements with a large stringer debond. The defects were caused by impacts performed from the backside of the plate. As result, the stringer feet debonds locally which was verified with conventional ultrasound measurements.
Detailliertes Wissen über die mechanischen Eigenschaften verwendeter Materialien ist Grundvoraussetzung für viele ingenieurtechnische Aufgaben und Dienstleistungen. Zur Bestimmung der elastischen Materialparameter gibt es verschiedene klassische, zerstörende Prüfverfahren. Eine Möglichkeit der zerstörungsfreien Bestimmung liegt in der Auswertung von Messergebnissen, die auf Basis des Ausbreitungsverhaltens geführter Ultraschallwellen gewonnen wurden. Das Ausbreitungsverhalten geführter Ultraschallwellen kann mittels Dispersionsabbildungen dargestellt werden.
Um aus messtechnisch ermittelten Dispersionsabbildungen Rückschlüsse auf die Materialparameter zu ziehen, werden in der aktuellen Forschung verschiedene inverse Methoden diskutiert. Maschinelles Lernen und insbesondere Convolutional Neural Networks (CNNs) stellen eine Möglichkeit der automatisierten inversen Modellierung und Evaluierung von Bilddaten dar.
In diesem Beitrag wird gezeigt, wie das Ausbreitungsverhalten von geführten Ultraschallwellen unter Verwendung von CNNs genutzt werden kann, um die isotropen elastischen Konstanten einer plattenförmigen Struktur zu bestimmen. Hierfür werden die verwendeten Daten analysiert, das Preprocessing erläutert und eine grundlegende CNN-Architektur gewählt. Zur Auswertung des generierten Modells werden verschiedene Verfahren wie Gradienten-Mapping und die Visualisierung der verschiedenen Schichten vorgestellt. Die Anwendbarkeit der Methode wird anhand synthetischer Daten demonstriert.
Composite-overwrapped pressure vessels (COPV) are increasingly used in the transportation industry due to their high strength to mass ratio. Throughout the years, various designs were developed and found their applications. Currently, there are five designs, which can be subdivided into two main categories - with a load-sharing metal liner and with a non-load-sharing plastic liner. The main damage mechanism defining the lifetime of the first type is fatigue of the metal liner, whereas for the second type it is fatigue of the composite overwrap. Nevertheless, one damage type which may drastically reduce the lifetime of COPV is impact-induced damage. Therefore, this barely visible damage needs to be assessed in a non-destructive way to decide whether the pressure vessel can be further used or has to be put out of service. One of the possible methods is based on ultrasonic waves. In this contribution, both conventional ultrasonic testing (UT) by high-frequency bulk waves and wavenumber mapping by low frequency guided waves are used to evaluate impact damage. Wavenumber mapping techniques are first benchmarked on a simulated aluminium panel then applied to experimental measurements acquired on a delaminated aluminium-CFRP composite plate which corresponds to a structure of COPV with a load-sharing metal liner. The analysis of experimental data obtained from measurements of guided waves propagating in an aluminium-CFRP composite plate with impact-induced damage is performed. All approaches show similar performance in terms of quantification of damage size and depths while being applied to numerical data. The approaches used on the experimental data deliver an accurate estimate of the in-plane size of the large delamination at the aluminium-CFRP interface but only a rough estimate of its depth. Moreover, none of the wavenumber mapping techniques used in the study can quantify every delamination between CFRP plies caused by the impact, which is the case for conventional UT. This may be solved by using higher frequencies (shorter wavelengths) or more advanced signal processing techniques. All in all, it can be concluded that imaging of complex impact damage in fibre-reinforced composites based on wavenumber mapping is not straightforward and stays a challenging task.
In the context of Industry 4.0 and especially in the field of Structural Health Monitoring, Condition Monitoring and Digital Twins, simulations are becoming more and more important. The exact determination of material parameters is required for realistic results of numerical simulations of the static and dynamic behavior of technical structures. There are many possibilities to determine
elastic material parameters. One possibility of non-destructive testing are ultrasonic guided waves. For the evaluation of the measurement results, mostly inverse methods are applied in order to be able to draw conclusions about the elastic material parameters from analysing the ultrasonic guided wave propagation. For the inverse determination of the elastic material Parameters with ultrasonic guided waves, several investigations were carried out, e.g. the determination of the isotropic material parameters through the point of zero-groupvelocity or anisotropic material parameters with a simplex algorithm. These investigations are based on the evaluation of dispersion images. Machine learning and in particular Convolutional Neural Networks (CNN) are one possibility of the automated evaluation from Image data, e.g. classification or object recognition problems. This article shows how the dispersive behavior of ultrasonic guided waves and CNNs can be used to determine the isotropic elastic constants of plate-like structures.
The Scaled Boundary Finite Element Method (SBFEM) is a semi-analytical method that shows promising results in modelling of guided ultrasonic waves. Efficiency and low computational cost of the method are achieved by a discretisation of the boundary of a computational domain only, whereas for the domain itself the analytical solution is used. By means of the SBFEM different types of defects, e.g. cracks, pores, delamination, corrosion, integrated into a structure consisting of anisotropic and isotropic materials can be modelled.
In this contribution, the SBFEM is used to analyse the propagation of guided waves in a structure consisting of an isotropic metal bonded to anisotropic carbon fibre reinforced material. The method allows appropriate wave types (modes) to be identified and to analyse their interaction with different defects. Results obtained are used to develop a structural health monitoring system for composite pressure vessels used in automotive and aerospace industries.
Die Scaled Boundary Finite Elemente Methode (SBFEM) ist eine semi-analytische Methode, die speziell für Modellierung von geführten Wellen weiterentwickelt und optimiert wurde. Da nur den Rand der Rechendomäne diskretisiert wird, hat die SBFEM einen geringen Rechenaufwand. In diesem Beitrag wird die SBFEM benutzt, um die Ausbreitung geführter Wellen in einer Metall-Faserverbund-Werkstoffstruktur zu analysieren. Mittels der SBFEM ist es möglich, verschiede Fehlertypen, z.B. Ermüdungsrisse, Poren, Delaminationen, Korrosion, in das numerische Modell zu integrieren und damit Defekt-Mode-Wechselwirkung zu analysieren. Die Ergebnisse wurden für die Entwicklung einer Methode zur Zustandsüberwachung von Composite-Druckbehältern verwendet.
Damit eine Simulationsrechnung, beispielsweise mit einer FEM-Software, eine ausreichend hohe Genauigkeit erreicht, muss vorausgesetzt werden, dass die Modellparameter eine sehr hohe Güte aufweisen. Die genaue Kenntnis der Materialparameter ist dabei von besonderer Bedeutung. Um diese Parameter bestimmen zu können, müssen die verwendeten Werkstoffe messtechnisch charakterisiert werden. Neben anderen Ansätzen sind dafür akustische Verfahren im Ultraschallbereich geeignet. Für dünnwandige und plattenförmige Materialien können aus den sich ausbreitenden geführten Wellen messtechnisch Dispersionskurven bestimmt und aus diesen die Materialparameter abgeleitet werden.
Da für die Signalverarbeitung und für Optimierungsaufgaben aktuell zunehmend Machine Learning Tools zum Einsatz kommen, stellt sich die Frage, ob diese Werkzeuge auch für die Ermittlung der Materialparameter aus den gemessenen Dispersionskurven eingesetzt werden können.
In der vorgestellten Untersuchung soll ein Convolutional Neural Network aufgestellt werden, welches aus Dispersionsbildern Muster extrahiert und aus diesen eine Schätzung für die Materialparameter ermittelt. Um die Machbarkeit dieses Ansatzes zu prüfen, werden zunächst nur isotrope Materialien betrachtet. Für das Netz werden mit der Scaled-Boundary-Finite-Element-Methode synthetische Daten für das Trainieren und Validieren generiert. Zusätzlich werden die Hyperparameter des neuronalen Netzes variiert, um ein optimales Model für die Schätzung zu finden. Anschließend kann das Netz mit experimentellen Daten getestet und das Ergebnis hinsichtlich der Genauigkeit bewertet werden.
Guided waves (GW) are of great interest for non-destructive testing (NDT) and structural health monitoring (SHM) of engineering structures such as for oil and gas pipelines, rails, aircraft components, adhesive bonds and possibly much more. Development of a technique based on GWs requires careful understanding obtained through modelling and analysis of wave propagation and mode-damage interaction due to the dispersion and multimodal character of GWs. The Scaled Boundary Finite Element Method (SBFEM) is a suitable numerical approach for this purpose allowing calculation of dispersion curves, mode shapes and GW propagation analysis. In this article, the SBFEM is used to analyse wave propagation in a plate consisting of an isotropic aluminium layer bonded as a hybrid to an anisotropic carbon fibre reinforced plastics layer. This hybrid Composite corresponds to one of those considered in a Type III composite pressure vessel used for storing gases, e.g., hydrogen in automotive and aerospace applications. The results show that most of the wave energy can be concentrated in a certain layer depending on the mode used, and by that damage present in this layer can be detected. The results obtained help to understand the wave propagation in multi-layered structures and are important for further development of NDT and SHM for Engineering structures consisting of multiple layers.