Cem Selcuk

An experimental study on the applicability of acoustic emission for wind turbine gearbox health diagnosis

Condition monitoring of wind turbine gearboxes has mainly relied upon vibration, oil analysis and temperature monitoring. However, these techniques are not well suited for detecting early stage damage. Acoustic emission is gaining ground as a complementary condition monitoring technique as it offers earlier fault detection capability compared with other more established techniques. The objective of early fault detection in wind turbine gearboxes is to avoid unexpected catastrophic breakdowns, thereby reducing maintenance costs and increase safety. The aim of this investigation is to present an experimental study the impact of operational conditions (load and torque) in the acoustic emission activity generated within the wind turbine gearbox. The acoustic emission signature for a healthy wind turbine gearbox was obtained as a function of torque and power output, for the full range of operational conditions. Envelope analysis was applied to the acoustic emission signals to investigate repetitive patterns and correlate them with specific gearbox components. The analysis methodology presented herewith can be used for the reliable assessment of wind turbine gearbox subcomponents using acoustic emission.

Increased range of ultrasonic guided wave testing of overhead transmission line cables using dispersion compensation

Overhead Transmission Line (OVTL) cables can experience structural defects and are, therefore, inspected using Non-Destructive Testing (NDT) techniques. Ultrasonic Guided Waves (UGW) is one NDT technique that has been investigated for inspection of these cables. For practical use, it is desirable to be able to inspect as long a section of cable as possible from a single location. This paper investigates increasing the UGW inspection range on Aluminium Conductor Steel Reinforced (ACSR) cables by compensating for dispersion using dispersion curve data. For ACSR cables, it was considered to be difficult to obtain accurate dispersion curves using modelling due to the complex geometry and unknown coupling between wire strands. Group velocity dispersion curves were, therefore, measured experimentally on an untensioned, 26.5m long cable and a method of calculating theoretical dispersion curves was obtained. Attenuation and dispersion compensation were then performed for a broadband Maximum Length Sequence (MLS) excitation signal. An increase in the Signal to Noise Ratio (SNR) of about 4-8dB compared to that of the dispersed signal was obtained. However, the main benefit was the increased ability to resolve the individual echoes from the end of the cable and an introduced defect in the form of a cut, which was 7 to at least 13dB greater than that of the dispersed signal. Five echoes were able to be clearly detected using MLS excitation signal, indicating the potential for an inspection range of up to 130m in each direction. To the best of the authors knowledge, this is the longest inspection range for ACSR cables reported in the literature, where typically cables, which were only one or two meter long, have been investigated previously. Narrow band tone burst and Hann windowed tone burst excitation signal also showed increased SNR and ability to resolve closely spaced echoes.

Coded Waveform Excitation for High-Resolution Ultrasonic Guided Wave Response

Ultrasonic guided wave-based nondestructive testing systems are widely used in various fields of industry where the structural integrity of components is of vital importance. Signal interpretation in these systems might become challenging due to multimodal and dispersive response of the interrogated structure. These phenomena degrade the signal-to-noise ratio and also lower the spatial/temporal resolution. This paper compares the use of Maximal Length Sequences and linear chirp excitation signals to develop a novel signal processing technique using dispersion compensation and cross-correlation. The technique is applied to both simulated and experimental multimodal signals from an aluminium rod for performance assessment. It is quantitatively validated that the technique noticeably improves the signal-to-noise ratio of the guided wave response and is able to acquire an accurate time of flight of the individual wave modes, and hence, the propagation distance. The technique is compared for both linear chirp and maximal length sequences excitation signals. Noise analysis for these excitation signals is also presented.

Pulse-compression based iterative time-of-flight extraction of dispersed Ultrasonic Guided Waves

Ultrasonic Guided Wave (UGW) based NonDestructive Testing (NDT) systems are widely used in numerous branches of industry, where the structural integrity of components carries vital importance. In those systems, signal interpretations might become challenging due to multi-modal and dispersive response of the structure under examination. This results in degradation of the signals in terms of Signal-to-Noise Ratio (SNR) and spatial/temporal resolution. This paper uses Maximal Length Sequences (MLS) to develop a novel signal processing technique by employing the Short-Time Fourier Transform (STFT), dispersion compensation and cross-correlation. The technique is applied to experimental multi-modal signals from an aluminum rod for performance verification. It is quantitatively validated that the technique noticeably improves the SNR of the guided wave response, and is able to derive an accurate time of flight of the individual wave modes and thus the propagation distance.

Developing a Novel Ice Protection System for Wind Turbine Blades Using Vibrations of Both Short and Long Wavelengths

Icing conditions in cold regions of the world may cause problems for wind turbine operations, since accreted ice can reduce the efficiency of power generation and create concerns regarding ice-shedding. This paper covers modelling studies and some experimental development for an ongoing ice protection system that provides both deicing and anti-icing actions for wind turbine blades. The modelling process contained two main sections. The first part involved simulation of vibrations with very short wavelength or ultrasonic guided waves (UGW) on the blade to determine optimal excitation frequency and transducer configuration. This excitation creates horizontal shear stress at the interface between ice and blade and focuses energy at the leading edge for de-bonding ice layers. The second modelling approach simulated the effects of vibrations with very long wavelength along with estimation of fatigue life due to harmonic forces to characterise the best parameters for shaker(s) mounted on blades. In parallel with this study, an empirical array of novel resonating shear transducers has been developed using a Design of Experiments (DoE) approach to demonstrate the practicability of inducing shear horizontal waves at the leading edge of wind turbine blades. This experimental verification also makes it possible to investigate the many parameters influencing ice-removal. In addition, piezo-electric and macro-fibre composite actuators have been investigated in place of conventional electromagnetic shakers, in order to save weight and simplify integration of the deicing system components. The ongoing research is intended to provide an active solution for icing prevention and deicing, enabling safe and reliable operation of wind turbines in adverse weather conditions.

Defect detection and classification system for automatic analysis of digital radiography images of PM parts

Abstract Digital radiography is a promising non-destructive testing tool for powder metallurgy (PM) parts, in which transmitted X-rays are recorded to generate data for an advanced defect detection system. An important part of this system is the data processing platform for pattern recognition in X-ray images. Combinations of advanced techniques for noise reduction, contrast enhancement and image segmentation are employed. Algorithms of registration for images in regions of interest are discussed, e.g. the scale invariant feature transform (SIFT). Modern pattern recognition methodologies such as smoothing, moment representation, image alignment and optical flow towards feature classification are evaluated. The proposed defect detection and classification capability for automatic analysis of digital radiographic images from PM parts potentially allows integration into multiple-view inspection systems, which should enhance quality control in the PM manufacturing and production environment. Defect detection systems able to work at the speed of current production lines are of great interest to both PM manufacturers and users.

Digital radiographic inspection technique for production friendly quality assessment of PM parts

Abstract Progress with the EU DIRA-Green/AutoInspect programme is reported. The target of the programme is to develop the capability for inspection of powder metallurgy (PM) parts – including sintered parts, ideally inline – in production environments, to improve the quality of output batches and to reduce scrap to the greatest possible extent. An important component of the inspection system, which is based on digital radiography, is the development of an advanced data processing system for pattern recognition and feature characterisation. This system will utilise modern image processing and pattern recognition techniques such as the Canny algorithm, active shape modelling and comparison with traditional cross-correlation and distance template matching. It is anticipated that the developed quality control system will be sufficiently versatile to favour its adoption by the wider PM community, among other industry sectors. In particular, the pattern recognition capability will be useful for integration into image processing and automation systems.

Assessment of Material Properties of Gallium Orthophosphate Piezoelectric Elements for Development of Phased Array Probes for Continuous Operation at 580°C

In this paper, the thickness extension mode gallium orthophosphate single crystal elements were characterised using the impedance analyser. Impedance characteristics of piezoelectric elements were investigated at temperatures from 25°C up to 580°C at first and then at a constant temperature of 580°C for a period of 25 days. The resonant and anti-resonant frequencies extracted from the impedance characteristics, capacitance (measured at 1 kHz), density and dimensions of the gallium orthophosphate elements were used to calculate electromechanical, piezoelectric and elastic properties of these elements at high temperatures as a function of time. The tested gallium orthophosphate elements proved to possess very stable efficiency and sensing capability when subjected to high temperature. The results are very encouraging for proceeding with development of phased array probes using gallium orthophosphate, for inspection and condition monitoring of high temperature pipelines in power plants at a temperature up to 580°C.

High Temperature Dielectric, Elastic and Piezoelectric Coefficients of Shear Type Lithium Niobate Crystals

In this paper shear type lithium niobate has been studied. The impedance-frequency characteristics were measured at high temperatures. This was carried out by placing the samples inside a furnace and performing in-situ impedance measurements at up to 600°C. The characteristic frequencies, capacitance, density and dimensions of samples were used to calculate the dielectric, elastic and piezoelectric coefficients. Prototype transducers were built using shear type lithium niobate crystals. The initial ultrasonic experiments show that it can be used at temperatures up to 450°C to transmit and receive guided wave signals at 70 kHz. This will enable design and manufacture of high temperature transducers for continuous guided wave monitoring of power plants.

Automated Quality Characterization for Composites Using Hybrid Ultrasonic Imaging Techniques

ABSTRACT An enhanced technique using image processing has been developed for automated ultrasonic inspection of composite materials, such as glass/carbon-fibre-reinforced polymer (GFRP or CFRP), to ascertain their structural healthiness. The proposed technique is capable of identifying the abnormality features buried in the composite by image filtering and segmentation applied to ultrasonic C-Scan images. This work presents results performed on two composite samples with simulated delamination defects. A local gating scheme is applied to raw A-Scan data for improved contrast between defective and healthy regions in the produced C-Scan image. In this test campaign, different filtering and thresholding algorithms are evaluated and compared in terms of their effectiveness on defect identification. The accuracies of less than 3 mm and 1.11 mm were attained for the defect size and depth, respectively. The results demonstrates the applicability of the proposed technique for accurate defect localization and characterization of composite materials.