Study On Damage Evolution In Wind Turbine Blades Based On Acoustic Emission Signal Processing

The blade is the key component of the wind turbine to obtain wind energy.In the production process of the blade,due to the special manufacturing process and the low degree of automation,the produced blade has internal defects such as wrinkles,delamination,and lack of glue.Due to the existence of these randomly distributed process defects,the fatigue failure of composite materials usually starts from the defects,and under the action of random alternating stress,it gradually expands and penetrates into a macro crack,and then gradually expands to the interface to cause fatigue damage,causing damage to the blade structure.Considering that most wind farms are located in remote areas,there are difficulties in maintenance and monitoring.If the early damage is not discovered in time,it may develop into a severe accident under severe conditions and cause huge economic losses.Therefore,it is of great significance to study the damage evolution identification of wind turbine blades to ensure the safe operation of the blades for a long time.This paper studies the use of acoustic emission technology to identify and predict the damage evolution status of blade composite materials,and provides new ideas for the monitoring of wind turbine blade health.The main research contents of the paper are as follows:(1)Based on the damage mechanics theory,by analyzing the energy dissipation of damage evolution at different stages,a damage evolution model of the wind turbine blade composite material was established,the relationship between the energy dissipation of acoustic emission and the damage evolution law of composite materials is clarified.The Lamb theory of composite laminates is used to discuss different types of Lamb dispersion control equations and dispersion characteristics.Using the acoustic emission lead-breaking experiment to analyze the different Lamb wave propagation modes and discuss the effects of different damage levels on the Lamb wave,which provides a theoretical basis for the analysis of the acoustic emission signal waveform during the damage evolution process.(2)According to the wrinkle and layering process defects that have the greatest impact on the blade quality in the wind turbine blade quality standards of the wind power generator set,a GFRP composite acoustic emission experimental platform was specifically established,and the experimental steps and artificial defect manufacturing methods were elaborated.The experiment analyzes the effects of layering defect location,size and fold defects with different aspect ratios on the mechanical properties of the composite.The cluster analysis algorithm was used to identify the damage patterns of composite materials,and the correctness of the damage pattern recognition was verified by electron microscope scanning.By analyzing the acoustic emission characteristics of composite materials with different defects,the influence of defect types and geometric parameters on the damage law of blades is clarified,which provides a basis for damage pattern recognition and condition monitoring of defective composite materials.(3)During the evolution of fold defects,due to the diversity of damage modes,the number of observed AE signal sources is smaller than that of acoustic emission sourc es,This paper proposes an improved K-means underdetermined blind source separation method,which effectively extracts the frequency characteristics of matrix cracking,fiber peeling,interface delamination and fiber fracture during the evolution of fold defect damage.Finally,the acoustic energy dissipation trend of various damage characteristics in fatigue damage evolution is calculated and analyzed.The research results show that the fold defect is the main source of energy dissipation in the longitudinal damage stage of the fiber and fiber bundles.The cracks and debonding and fiber breakage in the unstable damage stage are the main sources of AE excitation.And it shows the situation of high amplitude energy release,which clarifies the damage evolution mechanism of fold defects.(4)Considering the problem of cross-term interference caused by multi-component materials in the evolution of delamination defects of blade laminates,a time-frequency analysis method based on adaptive VMD-WVD is proposed,through the iterative search using the alternating direction method of multiplier to find the saddle point of the augmented Lagrange function,iteratively update the acoustic emission modal components and center frequency.Experimental results show that the performance of the algorithm is evaluated through the correlation coefficient and time-frequency resolution of the two-dimensional time-frequency distribution,this method effectively reduces cross-term interference,and can also ensure higher time-frequency aggregation and time-frequency resolution.It can meticulously describe the change process of the acoustic emission signal on the time-frequency plane,and characterize the complex dynamic process of layered defect damage.(5)It is considered that the blade composite material has developed into a macro failure before the stress reaches the maximum,which makes it difficult to identify and predict the instability damage.To solve this problem,a method for identifying and predicting precursor characteristics of composite material instability and destruction based on acoustic emission signal clustering analysis and neural network is proposed.By comparing the time-series evolution characteristics of each acoustic emission signal type,suitable precursor characteristic signals are screened out Establish a neural network prediction model.The results show that the method can effectively identify and predict its instability and destruction state compared with parameters such as accumulated energy and accumulated count of acoustic emission.