Blade Defect Research Articles

Wind turbine blade defect detection using hyperspectral imaging

Regardless of the evolution in the wind turbine industry, the operation of wind farms faces critical challenges when it comes to maintaining the lowest possible cost of energy. It is essential to early detect or predict wind turbine breakdowns due to different factors such as material degradation, electrical or mechanical failures, faults, or environmental damage. Wind turbine blades are the most expensive and most exposed parts of a wind turbine and suffer from many shortcomings, mainly cracks and erosion, which reduces their performance. Hence, there is an essential requirement for using non-destructive diagnostic of wind turbine blades. This paper lists some of the current non-destructive techniques for wind turbine blades analysis, their applicability, advantages, and drawbacks. Nevertheless, these methods face drawbacks that can be overcome by remote sensing. Hyperspectral imaging is a spectral imaging technique that integrates imaging and spectroscopy. It also enables the analysis and identification of distinctive spectral signatures and assigns them to the examined sample elements. Thus, this paper describes hyperspectral imaging implementation in image acquisition, handling, and flaw recognition as well as the detection of cracks and erosion. This technique's field output results show that blade defect detection's accuracy and precision are significantly enhanced.

Acoustic fields of unmanned aerial vehicles in the tasks of passive detection

The increasingly widespread use of unmanned aerial vehicles (UAVs) requires improved methods of their detection. In some cases, this is only possible by analyzing the physical fields created by the UAV. This work is dedicated to the study of the UAV’s acoustic field as applied to the problem of passive detection. In the laboratory, the acoustic fields of the most popular quadcopters were studied, as well as the properties of individual propellers of various designs. The influence of blade defects on the resulting spectral portrait was analyzed. Based on the results of numerical simulations in the COMSOL Multiphysics software package, directional patterns and pressure variations caused by a rotating propeller are obtained. A neural network has been trained to analyze the kepstrum coefficients calculated from acoustic records. The neural network allows monitoring and recognizing the presence of a quadrocopter in an area of interest, taking into account natural and anthropogenic noises. Field experiments carried out in urban conditions made it possible to estimate the range of possible passive detection of an unmanned aerial vehicle. [This work was financially supported by the Russian Foundation for Basic Research (Project No. 19-29-06062).]

Image Enhanced Mask R-CNN: A Deep Learning Pipeline with New Evaluation Measures for Wind Turbine Blade Defect Detection and Classification.

Demand for wind power has grown, and this has increased wind turbine blade (WTB) inspections and defect repairs. This paper empirically investigates the performance of state-of-the-art deep learning algorithms, namely, YOLOv3, YOLOv4, and Mask R-CNN for detecting and classifying defects by type. The paper proposes new performance evaluation measures suitable for defect detection tasks, and these are: Prediction Box Accuracy, Recognition Rate, and False Label Rate. Experiments were carried out using a dataset, provided by the industrial partner, that contains images from WTB inspections. Three variations of the dataset were constructed using different image augmentation settings. Results of the experiments revealed that on average, across all proposed evaluation measures, Mask R-CNN outperformed all other algorithms when transformation-based augmentations (i.e., rotation and flipping) were applied. In particular, when using the best dataset, the mean Weighted Average (mWA) values (i.e., mWA is the average of the proposed measures) achieved were: Mask R-CNN: 86.74%, YOLOv3: 70.08%, and YOLOv4: 78.28%. The paper also proposes a new defect detection pipeline, called Image Enhanced Mask R-CNN (IE Mask R-CNN), that includes the best combination of image enhancement and augmentation techniques for pre-processing the dataset, and a Mask R-CNN model tuned for the task of WTB defect detection and classification.

Review of a Custom-Designed Optical Sensing System for Aero-Engine Applications

Fibre bundle-based reflective optical sensors are good candidates for parameter monitorisation in aero engines. Tip clearance is one of those parameters of great concern that is necessary to monitor. Within this optical technology, the evolution experienced by a custom-designed optical sensor is presented from its first configuration up to the fifth one. The performance of the last configuration is compared with those of other two optical sensors that are also based on a fibre bundle design. The comparison has been carried out in an experimental program in a transonic wind tunnel for aero engines. The proven high resolution and sensitivity of the last configuration of the optical sensor opens up the possibility to detect blade defects, cracks, etc. that could otherwise be hard to track.

The application of ultrasonic phased array technology in wind turbine blade defect detection system

Start The working environment of wind turbines is complex and harsh. If the early damage of blades is not found in time, it may develop into mechanical failure under bad working conditions. Once an accident occurs, the consequences will be serious. In this paper, we study the non-destructive testing of ultrasonic phased array technique in the application of wind turbine blade defect detection, and analysis all kinds of common defects of blade production stage. According to these defects, the corresponding test device is selected, and the fault library is accumulated through defect test blocks. The original data of blade ultrasonic detection were identified, compared and finally the defect type was determined.

Image recognition of wind turbine blade damage based on a deep learning model with transfer learning and an ensemble learning classifier

An image recognition model based on a deep learning network is proposed for the automatic extraction of image features and the accurate and efficient detection of wind turbine blade damage. The Otsu threshold segmentation method is used to segment the blade image to eliminate the influence of the image background on the detection task. In order to improve the recognition performance of the proposed deep learning model, transfer learning and an ensemble learning classifier are used in a convolutional neural network model. Transfer learning is used to enhance the ability of the proposed model to extract abstract features and accelerate the convergence efficiency, whereas the random forest-based ensemble learning classifier is used to improve the accuracy of detecting the blade defects. The performance of the proposed model is verified by using unmanned aerial vehicle (UAV) images of the wind turbine blades. The proposed model provided better performance than the support vector machine (SVM) method, the basic deep learning model and the deep learning model combined with the ensemble learning approach.

Rotordynamic Analysis of a Dual-Spool Turbofan Engine with Focus on Blade Defect Events

Rotordynamic Analysis of a Dual-Spool Turbofan Engine with Focus on Blade Defect Events

항공기 가스터빈엔진 터빈블레이드의 고장률 예측에 관한 연구

The purpose of this study is to suggest a method for the efficient preventive maintenance of aircraft gas turbine engine turbine blades. For this study, the types and characteristics of gas turbine engines and its turbine blades were studied, the turbine blade defect types that caused an In-Flight Shut Down(IFSD) were analyzed, the blade failure rate according to the blade life cycle was analyzed through the Weibull distribution, one of the statistical techniques. Through these research results, it is possible to supplement the problems of the life cycle management and maintenance method of the turbine blade, and to suggest the measures to strengthen the preventive maintenance of the turbine blade. In this analysis, when total cycle of turbine blade exceeds 18,000 cycles, the failure rate is over 98%, and then the special management measures are required.

An end-to-end learning method for industrial defect detection

PurposeAutomatic defect detection is a fundamental and vital topic in the research field of industrial intelligence. In this work, the authors develop a more flexible deep learning method for the industrial defect detection.Design/methodology/approachThe authors propose a unified framework for detecting defects in industrial products or planar surfaces based on an end-to-end learning strategy. A lightweight deep learning architecture for blade defect detection is specifically demonstrated. In addition, a blade defect data set is collected with the dual-arm image collection system.FindingsNumerous experiments are conducted on the collected data set, and experimental results demonstrate that the proposed system can achieve satisfactory performance over other methods. Furthermore, the data equalization operation helps for a better defect detection result.Originality/valueAn end-to-end learning framework is established for defect detection. Although the adopted fully convolutional network has been extensively used for semantic segmentation in images, to the best knowledge of the authors, it has not been used for industrial defect detection. To remedy the difficulties of blade defect detection which has been analyzed above, the authors develop a new network architecture which integrates the residue learning to perform the efficient defect detection. A dual-arm data collection platform is constructed and extensive experimental validation are conducted.

Defect identification of wind turbine blades based on defect semantic features with transfer feature extractor

The monitoring of the status of the wind turbine blades is significant for the wind generation system and currently mainly dependent on manual visual inspections. The variance of the blade defects and the lack of the blade defect images make the defect identification of the wind turbine blades challenging. This paper proposes a defect identification method of wind turbine blades based on defect semantic features with transfer feature extractor. A deep convolutional neural network (DCNN) is built and is trained on the ImageNet Large Scale Visual Recognition Challenge dataset. The deep hierarchical features of the training blade images are extracted by the trained DCNN and fed into a classifier. By training on the labeled blade images, the first n layers of the trained DCNN is selected as the transfer feature extractor to extract the defect semantic features and the defect classifier is also obtained. The blade images can be diagnosed by the defect classifier based on the defect semantic features. The experiments are conducted on a real dataset of wind turbine blade images. The experimental results demonstrate the high learning ability of the proposed method from the small samples and its effectiveness for the defect identification of wind turbine blades.

Taxonomy of Gas Turbine Blade Defects

Context—The maintenance of aero engines is intricate, time-consuming, costly and has significant functional and safety implications. Engine blades and vanes are the most rejected parts during engine maintenance. Consequently, there is an ongoing need for more effective and efficient inspection processes. Purpose—This paper defines engine blade defects, assigns root-causes, shows causal links and cascade effects and provides a taxonomy system. Approach—Defect types were identified from the literature and maintenance manuals, categorisations were devised and an ontology was created. Results—Defect was categorised into Surface Damage, Wear, Material Separation and Material Deformation. A second categorisation identified potential causes of Impact, Environmental causes, Operational causes, Poor maintenance, Poor manufacturing and Fatigue. These two categorisations were integrated with an ontology. Originality—The work provides a single comprehensive illustrated list of engine blade defects, and a standardised defect terminology, which currently does not exist in the aviation industry. It proposes a taxonomy for both engine blade defects and root-causes, and shows that these may be related using an ontology.

Effects of leading edge defects on aerodynamic performance of the [formula omitted] airfoil

Defects at the leading edge of blades are an important source of power loss for wind turbines. In the present work, a systematic study is carried out on two typical types of defects, i.e., the surface concaved deep defects and surface distributed shallow defects on the S809 airfoil. Different defect shapes, ranges, equivalent depths and their influence mechanisms are investigated via CFD. For deep defects, an enclosed vortex is formed in the defect cavity, which suppresses the momentum exchange between the external flow and internal flow, so that the airfoil aerodynamic performance is highly sensitive to the defect opening range and is little affected by the defect shape and equivalent depth. Flow around the leading edge is strongly hindered by the deep defects and an elongated leading-edge separation bubble is formed at the suction side of the airfoil. At large angles of attack, flow at the suction side is dominated by flow separation at both the leading edge and trailing edge. In contrast, for shallow defects, all the defect equivalent depth, opening range and shape can significantly influence the airfoil aerodynamic performance; and quantitatively, the effect of defect equivalent depth is the most significant. For the present simulation cases with defects, the maximum lift coefficient is notably decreased (by 35% to 61%) accompanied by a sharp increase in drag coefficient (by 131% to 217%). Under dynamic pitching motions, the opening of the dynamic lift (drag)-coefficient hysteresis curve is effectively enlarged. The present work aims to provide an important reference for the maintenance and management of turbine blade defects.

Operational limits of blade coating associated with high aspect ratio pigments: Part II—cylindrical laboratory coater

Process issues such as blade deposits, scratches, spits, and defects on the coating surface develop during blade coating of paper at high solids content and fast machine speeds. Coatings that contain high aspect ratio pigments display these issues at lower solids concentrations and slower web speeds than spherical or block shaped pigments. The shape of the pigment particles influences the characteristics of the coating formulation after drying, and high aspect ratio pigments can lead to high gloss surfaces that have good printing properties and are of interest in barrier coatings. The rheology, particle packing, and water retention are all functions of the pigment shape and size distribution. The base sheet properties of porosity, surface roughness, and permeability interact with the coating and the blade coater operating conditions. The possible causes of poor runnability are part of an interrelated and complex system. A cylindrical laboratory coater (CLC) was utilized to determine the speed at which operational and quality issues begin at varying solids concentration for two pigments with different shape factors on two different paper surfaces with varying absorption rates. An operational window was developed for both pigments. Characteristics of the pigment suspensions were explored in order to find a relationship with the onset of operational problems. The high shear rate viscosity of these suspensions did not explain the operational windows. The difference between the solids concentration and the immobilization solids, which is a function of the shape factor, has a linear relationship to the runnability limits. This could be ascribed to the tight packing structure of the particles as the solids content approaches the immobilization solids.

Study on wind turbine blade defect detection system based on imaging array

Currently in the process of wind farm inspection, wind turbine blade appearance inspection mainly adopts the telescope or high-definition cameras, low detection efficiency, labor intensity and the precision is limited, in order to solve this problem, a kind of wind turbine blades defect recognition system based on image array is proposed. Through the joint of array camera and image processing server, the functions of the image acquisition, processing, and defect recognition and detection results output are implemented. The software of artificial intelligence deep learning based on neural network algorithm is used to identify the defects of blade image, and output quality analysis report, to realize automatic detection of wind turbine blade surface defect. The field measurement results show that the system greatly improves the efficiency and accuracy of wind turbine blade defect detection.

A summary of defect evolution and health monitoring of wind turbine blades

The wind turbine blade is one of most critical systems in terms of failure as the wind turbines are rapidly becoming larger and more complicated. Developing better design procedures are ongoing and the enhancements over the time are promising and requires reliable supportive monitoring and maintenance systems to ensure the satisfactory level of asset cost effectiveness. The purpose of the paper is to review the contributions concerning the wind turbine blade defects, in order to: understand how the defects in wind turbine blades evolve, i.e., initiate and propagate over the lifetime and to understand how the defects in wind turbine blades can be monitored. The paper reviews the mechanical dynamic models, testing and condition monitoring techniques. The paper proceeds to provide a generic defect evolution scenario with aim to enhance the fragmented understanding of the micro and macro-structure behaviour of the wind turbine blades.

A Blade Defect Diagnosis Method by Fusing Blade Tip Timing and Tip Clearance Information

Blade tip timing (BTT) technology is considered the most promising method for blade vibration measurements due to the advantages of its simplicity and non-contact measurement capacity. Nevertheless, BTT technology still suffers from two problems, which are (1) the requirements of domain expertise and prior knowledge of BTT signals analysis due to severe under-sampling; and (2) that the traditional BTT method can only judge whether there is a defect in the blade but it cannot judge the severity and the location of the defect. Thus, how to overcome the above drawbacks has become a big challenge. Aiming at under-sampled BTT signals, a feature learning method using a convolutional neural network (CNN) is introduced. In this way, some new fault-sensitive features can be adaptively learned from raw under-sampled data and it is therefore no longer necessary to rely on prior knowledge. At the same time, research has found that tip clearance (TC) is also very sensitive to the blade state, especially regarding defect severity and location. A novel analysis method fusing TC and BTT signals is proposed in this paper. The goal of this approach is to integrate tip clearance information with tip timing information for blade fault detection. The method consists of four key steps: First, we extract the TC and BTT signals from raw pulse data; second, TC statistical features and BTT deep learning features will be extracted and fused using the kernel principal component analysis (KPCA) method; then, model training and selection are carried out; and finally, 16 sets of experiments are carried out to validate the feasibility of the proposed method and the classification accuracy achieves 95%, which is far higher than the traditional diagnostic method.

Weak Defect Identification for Centrifugal Compressor Blade Crack Based on Pressure Sensors and Genetic Algorithm.

The Centrifugal compressor is a piece of key equipment for petrochemical factories. As the core component of a compressor, the blades suffer periodic vibration and flow induced excitation mechanism, which will lead to the occurrence of crack defect. Moreover, the induced blade defect usually has a serious impact on the normal operation of compressors and the safety of operators. Therefore, an effective blade crack identification method is particularly important for the reliable operation of compressors. Conventional non-destructive testing and evaluation (NDT&E) methods can detect the blade defect effectively, however, the compressors should shut down during the testing process which is time-consuming and costly. In addition, it can be known these methods are not suitable for the long-term on-line condition monitoring and cannot identify the blade defect in time. Therefore, the effective on-line condition monitoring and weak defect identification method should be further studied and proposed. Considering the blade vibration information is difficult to measure directly, pressure sensors mounted on the casing are used to sample airflow pressure pulsation signal on-line near the rotating impeller for the purpose of monitoring the blade condition indirectly in this paper. A big problem is that the blade abnormal vibration amplitude induced by the crack is always small and this feature information will be much weaker in the pressure signal. Therefore, it is usually difficult to identify blade defect characteristic frequency embedded in pressure pulsation signal by general signal processing methods due to the weakness of the feature information and the interference of strong noise. In this paper, continuous wavelet transform (CWT) is used to pre-process the sampled signal first. Then, the method of bistable stochastic resonance (SR) based on Woods-Saxon and Gaussian (WSG) potential is applied to enhance the weak characteristic frequency contained in the pressure pulsation signal. Genetic algorithm (GA) is used to obtain optimal parameters for this SR system to improve its feature enhancement performance. The analysis result of experimental signal shows the validity of the proposed method for the enhancement and identification of weak defect characteristic. In the end, strain test is carried out to further verify the accuracy and reliability of the analysis result obtained by pressure pulsation signal.

표면파 기반의 풍력발전기 블레이드 표면상태 실시간 모니터링에 관한 연구

현재 풍력발전은 우리나라에서 가장 주목 받고 있는 신재생에너지 분야로 많은 연구가 진행 중이다. 풍력발전을 구성하는 요소 중 핵심 요소인 블레이드에 손상이 발생할 경우에 발전 효율에 직접적인 영향을 미치므로 효율적인 유지보수를 위해 초기에 결함을 측정하는 기술이 매우 중요한 상태이다. 그러나 기존의 초음파 비파괴 검사 및 열화상 비파괴 검사는 소요시간이 길고 실시간 모니터링이 어려우므로 초기 결함 측정이 불가능하다. 기존의 문제를 보완하고자 본 논문에서는 표면파를 이용한 블레이드 표면상태 실시간 모니터링에 관한 연구를 수행하였다. 압전센서 기반의 시스템을 구성하여 블레이드 표면샘플에 대해 공정 변수 별 기초 성능 실험을 하였고, 소형 블레이드에 대해 공정변수 별 실험을 통해 블레이드의 크랙, 벗겨짐, 장애물 등을 실시간으로 모니터링이 가능한지 연구 하였다. These days much efforts are being dedicated to wind power as a potential source of renewable energy. To maintain effective and uniform generation of energy, defect preservation of turbine blade is essential because it directly takes effects on the efficiency of power generation. For the effective maintenance, early measurements of blade defects are very important. However, current technologies such as ultrasonic waves and thermal imaging inspection methods are not suitable because of long inspection time and non-real time inspection. To supplement the problems, the study introduced a method for real time defect monitoring of a blade surface based on surface wave technology. We examined the effect of various parameters such as micro-cracks and peelings on the propagation of surface wave.

Band Saw Blade Crack before and after Comparison and Analysis of Experiments (2)

Based on MJ3310 woodworking band saw machine as the research object, under the no-load and load of Vib system vibration signal acquisition, processing and analysis software of band saw blade transverse vibration test and the signal acquisition and analysis of the collected signals obtained: to determine the transverse vibration displacement 5.66μm ~ 7.86μm and the main vibration frequency between 624 Hz ~ 792 Hz, then saw blade crack at least 3 mm, need timely saw blade, cutting high hardness of wood band saw blade transverse vibration displacement and frequency will increase sharply. Can be generated according to the band saw blade crack before and after the changing rule of the horizontal vibration displacement and frequency of transverse vibration and scope, judgment and replacement time of saw blade saw blade defect types, which can fully rational utilization of saw blade work effectively.

In Situ Eddy-Current Testing on Low-Pressure Turbine Blades of Aircraft Engine

Abstract To realize in situ testing on low-pressure turbine blades of aircraft engines, eddy-current testing was studied through probe design, artificial defect sample testing, and practical field applications. The eddy-current probe was composed of a differential coil with dual magnetic cores, flexible spring steel strip, and elongating composite shank, which, respectively, provided high detection sensitivity on the turbine blade defect, appropriate contact between the probe and the curving blade surface, and adequate extent for the long distance of in situ inspection. Practical applications in engine turbine blades testing proved that the eddy-current probe successfully solved the in situ testing difficulties, including poor accessibility, limited detection space, and strong interference signals.