Turbine Blade Surface Research Articles

Lightweight algorithm based on you only look once version 5 for multiple class defect detection on wind turbine blade surfaces

A lightweight wind turbine blade surface multiple class defect detection algorithm based on You Only Look Once version 5 (YOLOv5) is proposed to address the issues of low accuracy, poor efficiency, and high power consumption associated with manual and traditional automation algorithms used for detecting surface defects on wind turbine blades. Firstly, by replacing the Concentrated-Comprehensive Convolution Block module in YOLOv5 with the GhostBottleneckv2 module, the overall parameter count of the model is significantly reduced. Additionally, advocating the inclusion of the Convolutional Block Attention Module (CBAM) into the YOLOv5 backbone network to address the issue of small targets and significant scale variations in surface defects found on wind turbine blades. The YOLOv5 feature fusion module replaces the Feature Pyramid Network (FPN) structure with the Bidirectional Feature Pyramid Network (BiFPN) structure to enhance the fusion capability of multi-scale weighted features. Finally, the optimized model files underwent structured pruning through the application of pruning strategies, retraining fine-tuning, and knowledge distillation techniques. The experimental results on the self-made dataset of surface defects on wind turbine blades indicate that the optimized model has a parameter count that is only 48.79% of the small version of the YOLOv5 model. However, its mean Average Precision (mAP) and frame rate have increased by 5.05% and 8.1 frames/s, respectively, reaching 94.56% and 43.7 frames/s. The lightweight algorithm proposed in this study meets the accuracy and real-time requirements for surface defect detection on wind turbine blades.

Study on the influence of sediment erosion of Francis turbine runner under different water heads

Abstract The average measured sediment concentration of the river section where Wanjiazhai Hydropower Station is located over the years is 5.7kg/m 3, The maximum measured sediment concentration reaches 37.6kg/m 3, The erosion of the flow components of hydraulic turbines caused by sand containing water flow is becoming increasingly severe, coupled with the frequent impact of load changes during peak shaving operation, the problems exposed during the operation of hydraulic turbines are becoming more prominent, which affects the stable operation of the units. This article uses numerical simulation and experimental testing to study the sediment wear characteristics and hydraulic performance of the water turbine at Wanjiazhai Hydropower Station. Research has revealed the effects of working head on impeller wear. The increase in working head will exacerbate the impact wear on the main channel wall of the unit. The surface of water turbine blades is the main wear area, and the high-speed impact of particles causes an increase in the wall wear rate. The probability of particles coming into contact with the wall at the outlet increases, resulting in rapid wear at this point. The increase in head will intensify the wear intensity of the unit surface, and the flow velocity in the gap indicates that the local high-speed jet in the gap is the main cause of gap wear. Therefore, in order to ensure the efficient anti-wear performance of the water turbine, multiple factors need to be comprehensively considered to cope with sediment wear of the water turbine, including regular cleaning, optimized structural design, use of wear-resistant materials, replacement of worn parts, and optimization of operating parameters. These measures help to reduce the damage of sediment wear to water turbines and extend their service life.

Large eddy simulation of in-hole blockage effects on cooling performance of fan-shaped holes

Film-cooling and thermal barrier coating approaches are commonly used in gas turbines to protect the turbine parts from the high-temperature flow of gases. However, one drawback of this method is that the coatings can partially obstruct the cooling hole exit on the turbine blade surfaces, resulting in reduced cooling performance. Therefore, this study aimed to investigate the effects of in-hole blockage thickness on cooling effectiveness and flow characteristics of fan-shaped holes using large eddy simulations (LES). A laidback fan-shaped hole located on a flat plate was the reference cooling hole. Numerical simulations were conducted for two different blocked cooling hole configurations with various blockage thicknesses (t/D = 0.5 and 0.9) at three different blowing ratios (0.5, 1.0, and 2.0). The cooling effectiveness computed by the LES method was validated for both unblocked and blocked cooling holes, and the results were compared to the experiments at various blowing ratios. It was revealed that an increase in blockage thickness had a significant impact on the area-averaged cooling effectiveness, particularly at high blowing ratios. The overall area-averaged cooling effectiveness for the t/D = 0.9 case was approximately 75 % lower than that of the unblocked cooling hole. Moreover, assessment of the instantaneous velocity field inside the hole over time revealed that larger blockage thicknesses resulted in greater velocity fluctuations and a more unsteady flow.

IMPACT OF THE WALL-PARALLEL LENGTH SCALE OF ROUGHNESS ON THE DEVELOPMENT OF TURBULENT BOUNDARY LAYERS IN TURBINES

Abstract The detrimental effects of surface roughness on turbulent boundary layers in turbines, and compressors, are well known. Prediction can be problematic, especially for surfaces which are not similar to sand grains. Several publications have proposed additional parameters, slope, skew, etc. to augment a wallnormal measure. Here, we introduce a new roughness parameter, the mean separation, which explicitly measures the average separation between each local minima and its closest local maxima. Scans of turbine blade surfaces have mean separations which are up to six times that of the sand grain surface with the same wallnormal measure. High resolution, < 0.1 μm, area scanning tools and 3-D printing techniques have enabled the development of a “Scan-Scale-Print-Measure” methodology. A surface scan is scaled-up, 3-D printed, applied to a flat-plate and then the turbulent boundary layer is measured. The methodology has enabled parametric studies where the mean separation was increased whilst keeping either the wallnormal or the feature size constant. Both studies demonstrated a surface roughness-loss that was strongly dependent on the mean separation and different to a sand grain correlation. It also showed that for large mean separations the roughness-loss decreased to zero. A study into pits-and-peaks showed that, at the same wallnormal, peaks generated a roughness-loss twice that of pits. An engine representative surface produced only half the roughness-loss that would be attributed to a sand grain surface. The roughness loss was estimated within 5% using the parametric studies.

A pyramid-style neural network model with alterable input for reconstruction of physics field on turbine blade surface from various sparse measurements

Data from turbine cascade experiments typically exhibits low spatial–temporal resolution, along with inevitable noise and local data missing. This paper aims to establish a super-resolution model to reconstruct the complete pressure and temperature fields on the blade surface from limited observations, using the deep learning method. Depending on the three-dimensional geometric complexity of the blade, the required cross-sectional data varies, resulting in input data of different sizes. Conventional surrogate models typically confine themselves to a single input format, leading to limited compatibility with diverse inputs. A pyramid-style model with four optional input ports is proposed, designed to accommodate data with different numbers of span-wise sections. Three training strategies have been evaluated, where distributed training has been proven to be more time-effective and flexible while maintaining high prediction precision, compared to holistic and conventional training. Generally, the average relative error over the whole dataset falls below 0.18%, the average peak signal-to-noise ratio exceeds 52 dB, and the average structural similarity index measurement surpasses 0.999. As the number of span-wise sections and the amount of information in the input increase, the overall performance shows a consistent upward trend. Robustness analyses are conducted by applying artificial noises of varying intensities. The results indicate that the model can tolerate noise intensities of up to 5% with a satisfactory reconstruction accuracy. In addition, the model is quite sensitive to the measurement data noises in complex flow regions such as expansion waves, suggesting that more intensive measurements should be targeted to those regions when conducting cascade experiments. The proposed method satisfies the intended objectives and provides an idea for future applications in digital twin platforms.

Attention mechanism based on deep learning for defect detection of wind turbine blade via multi-scale features

Abstract An enhanced wind turbine blade surface defect detection algorithm, CGIW-YOLOv8, has been introduced to tackle the problems of uneven distribution of defect samples, confusion between defects and background, and variations in target scales that arise during drone maintenance of wind turbine blades. This algorithm is given based on the YOLOv8 model. Initially, a data augmentation method based on geometric changes and Poisson mixing was used to enrich the dataset and address the problem of uneven sample distribution. Subsequently, the incorporation of the Coordinate Attention (CA) mechanism into the Backbone network improved the feature extraction capability in complex backgrounds. In the Neck, the Reparameterized Generalized Feature Pyramid Network (Rep-GFPN) was introduced as a path fusion strategy and multiple cross-scale connections are fused, which effectively enhances the multi-scale expression ability of the network. Finally, the original CIOU loss function was replaced with Inner-WIoU, which was created by applying the Inner-IoU loss function to the Wise-IoU loss function. It improved detection accuracy while simultaneously speeding up the model's rate of convergence. Experimental results show that the mAP of the method for defect detection reaches 92%, which is 5.5% higher than the baseline network. The detection speed is 120.5 FPS, which meets the needs of real-time detection.

Numerical Study on the Impact Pressure of Droplets on Wind Turbine Blades Using a Whirling Arm Rain Erosion Tester

The leading-edge erosion of a wind turbine blade was tested using a whirling arm rain erosion tester, whose rotation rate is considerably higher than that of a full-scale wind turbine owing to the scale effect. In this study, we assessed the impact pressure of droplets on a wet surface of wind turbine blades using numerical simulation of liquid droplet impact by solving the Navier–Stokes equations combined with the volume-of-fluid method. This was conducted in combination with an estimation of liquid film thickness on the rotating blade using an approximate solution of Navier–Stokes equations considering the centrifugal and Coriolis forces. Our study revealed that the impact pressure on the rain erosion tester exceeded that on the wind turbine blade, attributed to the thinner liquid film on the rain erosion tester than on the wind turbine blade caused by the influence of centrifugal and Coriolis forces. This indicates the importance of correcting the influence of liquid-film thickness in estimating the impact velocity of droplets on the wind turbine blade. Furthermore, we demonstrated the correction procedure when estimating the impact velocity of droplets on the wind turbine blade.

Efficient Anti-Icing of a Stable PFA Coating for Wind Power Turbine Blades.

Superhydrophobic coatings are increasingly recognized as a promising approach to enhancing power generation efficiency and prolonging the operational lifespan of wind turbines. In this research, a durable superhydrophobic perfluoroalkoxy alkane (PFA) coating was developed and specifically designed for spray application onto the surface of wind turbine blades. The PFA coating features a micronano hierarchical structure, exhibiting a high water contact angle of 167.0° and a low sliding angle of 1.7°. The optimal PFA coating exhibits stability and maintains a superhydrophobic performance during mechanical and chemical tests. The findings of this study establish a positive association between the surface energy of the coating and its effectiveness in anti-icing. The delayed icing time for the PFA-coated surface is 46.83 times longer than that of an uncoated surface, and the ice adhesion strength is only 1.875 kPa. Additionally, the PFA coating demonstrates remarkably high ice suppression efficiencies of 94.7 and 99.5% in anti-icing experiments at ambient temperatures of -6 and -10 °C, respectively. It is anticipated that this stable superhydrophobic PFA coating will be a candidate for anti-icing applications in wind turbine blades.

Wind turbine blade defect detection with a semi-supervised deep learning framework

To increase the economic efficiency of utilizing wind turbines, an enhanced object detection model based on the small version of the fifth generation of the You Look Only Once algorithm (YOLOv5s) is proposed in this paper, which can effectively detect cracks on the surface of wind turbine blades using images captured by unmanned aerial vehicles. To improve the extraction of light-colored and low-definition images, Omni-Dimensional Dynamic Convolution (ODConv) and Dynamic Head Module components (DyHead) are introduced in the proposed method, while a lightweight Group Shuffle convolution (GSConv) module is utilized to accelerate the model inference speed without sacrificing detection performance. Furthermore, a semi-supervised learning strategy is developed to reduce human labors in annotating images. Extensive experiments demonstrate that the proposed model outperforms the original YOLOv5s in terms of both detection accuracy and inference speed. Besides, the proposed model has good performance against state-of-the-art methods. Furthermore, the experiments validate the efficacy of the proposed semi-supervised learning strategy.

An adaptive detection approach for multi-scale defects on wind turbine blade surface

Multi-scale defects will inevitably appear in wind turbine blade defect detection within practical detection applications. An adaptive multi-scale detection approach is proposed to accurately classify and locate defects on the wind turbine blade surface. The proposed approach includes two main detection procedures: rough detection and precise detection. In the rough detection, the multi-level features extraction module with an adaptive bounding box proposal module is used to depict multi-scale defect regions and train a binary classifier to distinguish defects from non-defects. At the precise detection stage, three defect categories and four coordinates representing defect locations are obtained based on a multi-class defect classifier and regression of rough location boxes. The proposed method is evaluated on a real wind turbine blade surface defect dataset collected in a commercial wind farm and annotated manually. Results show that (1) the proposed model can detect the class of the blade multi-scale defects and outperforms other schemes with 96.89% mAP in the same model training epochs, (2) the positioning performance analysis of the model for multi-scale defects is conducted to validate the accuracy of the proposed model for multi-scale defect location.

3D printing of thick film NTC thermistor from preceramic polymer composites for ultra-high temperature measurement

Integrating thick/thin film sensors into component systems has emerged as a prevalent approach for monitoring in extreme environments. However, traditional vapor deposition methods face obstacles, including complex fabrication processes and the degradation of sensitive materials at extremely high temperatures. This work delineates the development of a polysilazane composite dual-layer thick-film Negative Temperature Coefficient (NTC) thermistor characterized by its suitability for extreme temperatures and robust bond strength achieved through an advanced near-net-shape printing methodology. High-temperature resistant La(Ca)CrO3/polysilazane films were printed as the sensitive layer, while a dense layer formed by Cr2O3/polysilazane was used as the protective layer. The bilayer structure resulted in a 2.5-fold increase in adhesion strength compared to the single-layer La(Ca)CrO3/polysilazane films. Experimental results indicate that the dual-layer thick-film NTC thermistor can be operated long-term at 1300 °C with a resistance drift rate of 0.9 %/h and survive short-term exposure to temperatures up to 1550 °C. As a proof of concept, this work applied 3D printing technology to fabricate a polysilazane composite dual-layer thick-film NTC thermistor on the surface of turbine blades and demonstrated its functionality under flame impingement at nearly 1300 °C. Such flexible 3D printing techniques pave the way for a new paradigm in manufacturing sensors capable of withstanding ultra-high temperatures.

Research on formation mechanism and output effect of wind turbine ice-covered blades

Considering the physical characteristics of wind turbine wing icing, icing synthesis rate, and icing type, we selected the icing type and surface roughness of ice-coated blades as sensitive parameters. The focus of our research was on the equivalent particle roughness height correction model, and we numerically analyzed the two icing processes (frost ice and clear ice) on wind turbine blade surfaces by combining FENSAP-ICE and FLUENT analysis tools. We predicted the ice type on blade surfaces using a multi-time step method and analyzed how variations in icing shape and ice surface roughness affect the aerodynamic performance of blades during frost ice formation or clear ice formation. Our results indicate that differences in blade surface roughness and heat flux lead to disparities in both ice formation rate and shape between frost ice and clear ice. Clear ice has a greater impact on aerodynamics compared to frost ice, while frost ice is significantly influenced by the roughness of its icy surface. These findings can serve as valuable references for wind power operators and manufacturers seeking solutions to issues related to blade surface icing under extremely cold conditions.

Research on Thermocouple Integration Method of Hollow Thin-Walled Turbine Rotor Blade Surface

The hollow thin-walled structure is increasingly used in advanced aero-engine turbine rotor blades, and the traditional surface temperature measurement method based on groove buried couple is no longer applicable. In this paper, a new thermocouple integration method is proposed. The laser additive manufacturing process is used to generate thermocouple embedding channels on the surface of turbine blades. After the thermocouple is embedded, it is pre-fixed, and the thermocouple protection is realized by supersonic flame spraying technology. The strength of the thermocouple integrated structure of the turbine blade is verified by finite element analysis. Based on CFD analysis, the influence of thermocouple integration on the efficiency, pressure ratio and temperature distribution of the turbine blade is studied. The high-speed rotation test shows that the thermocouple integrated structure has high strength and reliability and meets the measurement requirements. This method can provide technical support for the surface temperature measurement of hollow thin-walled turbine rotor blades.

Automatic Defect Detection of Jet Engine Turbine and Compressor Blade Surface Coatings Using a Deep Learning-Based Algorithm

The application of additive manufacturing (AM) in the aerospace industry has led to the production of very complex parts like jet engine components, including turbine and compressor blades, that are difficult to manufacture using any other conventional manufacturing process but can be manufactured using the AM process. However, defects like nicks, surface irregularities, and edge imperfections can arise during the production process, potentivally affecting the operational integrity and safety of jet engines. Aiming at the problems of poor accuracy and below-standard efficiency in existing methodologies, this study introduces a deep learning approach using the You Only Look Once version 8 (YOLOv8) algorithm to detect surface, nick, and edge defects on jet engine turbine and compressor blades. The proposed method achieves high accuracy and speed, making it a practical solution for detecting surface defects in AM turbine and compressor blade specimens, particularly in the context of quality control and surface treatment processes in AM. The experimental findings confirmed that, in comparison to earlier automatic defect recognition procedures, the YOLOv8 model effectively detected nicks, edge defects, and surface defects in the turbine and compressor blade dataset, attaining an elevated level of accuracy in defect detection, reaching up to 99.5% in just 280 s.

A new stacking model method to solve an inverse flow and heat coupling problem for aero-engine turbine blades

The whole temperature distribution of a turbine blade is an important information for heat transfer design for turbine, but it is very difficult and expensive to measure the whole temperature filed, especially for the internal points of a blade. It is necessary to reconstruct the whole temperature distribution by calculating reversely unknown parameters based on sparse temperature values on the blade surface. This paper proposes a new stacking model method for solving inverse heat transfer problems of boundary conditions and it can be applied to cope with multi-input and multi-output regression problems. In order to improve the prediction performance of the stacking model, the Random Forest Regression, the XGBoost Regression and the K-Nearest Neighbors Regression are selected as the base estimators, and the Linear Regression is selected as the meta estimator. An application example for an aero-engine turbine blade is modeled and solved, in which the boundary condition parameters of the exhaust flows are reversely calculated and temperature distributions of a turbine blade surface can be obtained. Numerical experiments have demonstrated that the stacking model outperforms individual base models. This research is of considerable significance to heat transfer design, experimental research and decision support for real-time remote monitoring of aero-engines.

Piezoelectric-pneumatic material jetting printing for non-contact conformal fabrication of high-temperature thick-film sensors

Conformal printing of high-temperature thick-film sensors (TFSs) on intricate surfaces is crucial for precise integration and enhanced functionality. However, traditional contact printing faces challenges in terms of the complex motion control of 3D surfaces. With the core concept of non-contact printing, this study employs piezoelectric-pneumatic material jetting (PPMJ) to achieve rapid and uniform manufacturing of complex and multi-material conformal TFSs. PPMJ enables a wide viscosity range (50–100,000 mPa.s) and high solid content (60 wt%), accommodating various types of inks, from insulating to semiconductor materials to high-temperature alloy slurries, thus providing a broad operational window for shape control and performance modulation in the conformal manufacturing of TFSs on curved surfaces. Microscale shape control of ceramic ink features, including a parallel three-phase contact line, uniform print dimensions, and post-curing shape control, was systematically investigated. PPMJ printing facilitates the successful in situ fabrication of multi-layer TFSs, featuring the self-assembly of droplets and dependable printing consistency, with nozzle-to-substrate heights ranging from 1 to 10 mm, thus advancing the structural arrangement, design, and manufacturing of conformal multi-layer TFSs on curved surfaces. For example, the polymer-derived ceramic heat flux sensor printed by PPMJ demonstrates repeatable high-temperature responsiveness and can be integrated on the surface of turbine blades, withstanding thermal impacts from flames exceeding 800 °C. Hence, PPMJ printing technology opens new possibilities for the conformal manufacturing of various sensors on complex 3D surfaces.

Separation-induced transition on a T106A blade under low and elevated free stream turbulence

The separation-induced transition on the suction surface of a T106A low pressure turbine blade is a complex phenomenon with implications for aerodynamic performance. In this numerical investigation, we explore an adverse pressure gradient-dominated flow subjected to varying levels of free stream excitation, as the underlying separation-induced transition is a critical factor in assessing blade profile loss. By comprehensively analyzing the effects of free stream turbulence (FST) on the transition process, we delve into the various mechanisms which govern the instabilities underlying bypass transition by studying the instantaneous enstrophy field. This involves solving the two-dimensional (2D) compressible Navier–Stokes equation through a series of numerical simulations, comparing a baseline flow to cases where FST with varying turbulent intensity (Tu=4% and 7%) is imposed at the inflow. Consistent with previous studies, the introduction of FST is observed to delay flow separation and trigger early transition. We explore the different stages of bypass transition, from the initial growth of disturbances (described by linear stability theory) to the emergence of unsteady separation bubbles that merge into turbulent spots (due to nonlinear interactions), by examining the vorticity dynamics. Utilizing the compressible enstrophy transport equation for the flow in a T106A blade passage, we highlight the various routes of bypass transition resulting from different levels of FST, emphasizing the relative contributions from baroclinicity, compressibility, and viscous terms.

A review of icing prediction techniques for four typical surfaces in low-temperature natural environments

As a common phenomenon in nature and industrial fields, icing always adversely affects life, and plays negative effects. Although series of anti-/de-icing techniques are widely investigated, icing prediction techniques on surfaces under low-temperature natural environment are more important due to effectively reduce or even prevent icing caused harms. To understand the current research progress of icing prediction techniques, four typical static and moving surfaces that are prone to ice formation were selected as the research objects in this study, including road, transmission line, wind turbine blade and aircraft surfaces. As summarized, prediction methods mainly include physical models, statistical models, and machine learning. For road surface, statistical analysis, theoretical analysis and data mining methods are widely used, with the average prediction accuracy reaching 80%. For wind turbine blade and transmission line surfaces, both model-driven and data-driven methods were used, resulting in average prediction accuracies of 80% and 90%, respectively. For aircraft surface, the method of numerical analysis combined with machine learning is the mainstay, with a deviation of less than 20%. Summary and outlook are finally given, which are beneficial for the optimization of ice prediction technology in different industrial fields.

Preparation and performance study of SiO2 aerogel thermal insulation coating with nanoporous structure for wind turbine blade surface

To prepare a base layer of thermal insulation for the surface of wind turbine blades, we utilized SiO2 aerogel material with a nanoporous structure as the thermal insulation layer of an inorganic composite photothermal de-icing coating. We further investigated the influence of raw material ratio and process on the thermal insulation layer’s performance. By adjusting the process parameters, the microstructure and characteristics of the SiO2 aerogel thermal insulation coating were controlled. Microstructure scanning and EDS analysis were employed to assess the results. The results of the experiment suggest that, as opposed to brush coating, atomized spraying is a better technique for creating SiO2 aerogel thermal insulation coating. The coating possesses excellent mechanical stability and applicability, maintaining a coating wear rate within 5% after 20 wear cycles under a load of 1734 N/m2, and the coating achieves the highest class 0 surface adhesion to the substrate surface. The coating exhibits exceptional heat insulation performance, reaching a temperature of 32.7 °C after five spraying cycles, which is 20.0 ∼ 29.4 °C higher than the surface without producing a complete heat insulation layer. The SiO2 aerogel aqueous slurry has a viscosity of 31.3 ∼ 36.9 mPa∙s, which is capable of forming a uniform membrane surface when sprayed at a 60° angle at 0.95 MPa pressure. Consequently, this paper’s design of photothermal thermal insulation offers strong anti-abrasion and excellent thermal insulation, presenting a new avenue for scientific investigation into coating photothermal de-icing.

Method of scanning thin surfaces in repairing gas turbine blades

Improving the efficiency of weld overlay repair of gas turbine blades by developing and implementing a method of scanning complex curved surfaces of gas turbine blades directly on the machine is the aim of this work. The paper proposes an approach to scanning a blade with a computer vision system mounted near the nozzle on the same machine that is used for direct metal deposition. The computer vision system consists of a triangulation laser sensor and a camera. The proposed algorithm is adaptive to the mechanical condition of the equipment used for scanning and metal deposition. The scans obtained from the machine vision system have precision better than 0.05mm in 67.56% of cases, and precision better than 0.1mm in 95.75% of cases. That accuracy, with a laser spot of 0.5 to 1.0mm, is sufficient for further use of the scans in repairing gas turbine blades. The proposed approach makes it possible to speed up the preparation of technological programs for direct metal deposition by 10 times compared to the manual scanning method.