Study On Acoustic Signal Processing And Characteristics Based On No.45 Steel

As a dynamic nondestructive testing technique,acoustic emission technology can detect the early crack initiation of metal specimens in a more timely manner than other testing methods,define the damage degree,and provide the location of the damage.It is of great significance to the safe operation of large metal components and equipment.In this paper,acoustic emission test was conducted on no.45 steel specimen based on acoustic emission test technology,and acoustic signals of fatigue damage under cyclic loading were collected.A series of innovative research methods were proposed to study the acoustic signals at different damage stages.The main work of this paper is as follows:(1)By exploring the background and practical significance of acoustic emission technology,we deeply realize the application prospect of acoustic emission detection technology as a real-time dynamic detection technology,and summarize the development history and research status of acoustic emission technology at home and abroad.(2)Based on the basic principle of acoustic emission technology,characteristics of acoustic emission detection technology,basic characteristics of acoustic signal,basic types of acoustic signal and acoustic emission source,two important methods of acoustic signal processing and the formation reasons,characteristics and influencing factors of fatigue damage acoustic signal are analyzed.The experiment of fatigue damage acoustic signal acquisition under cyclic loading of no.45 steel specimen was carried out.Firstly,the required equipment and loading device are selected according to the characteristics of the specimen,and the parameters of relevant equipment and loading mode of the fatigue testing machine are set according to the experimental requirements.Under the guidance of the experimental scheme.acoustic signal acquisition experiments under cyclic loading were conducted.(3)A combined denoising method based on variational mode decomposition and singular value denoising is proposed,and wavelet threshold denoising and EMD denoising are used to compare the simulated and measured acoustic signals.The comparison results show that the new noise reduction method proposed in this paper has better noise reduction effect for acoustic signals under cyclic loading,both in terms of noise reduction effect diagram and evaluation index.(4)In this paper,time-frequency domain characteristics and IMF component under VMD decomposition are used to extract the features of the acoustic signal after noise reduction.The processing results show that the time-frequency domain features can not be directly extracted to see the damage stage of the sound signal.Therefore,this paper adopts the smoothing characteristic parameter processing method.The experimental results show that the damage stage can be clearly seen by smoothing time-frequency domain characteristics.Next,the feature extraction method combining wavelet packet and sample entropy was used to extract the features of prefabricated defective specimens under different loading conditions and fatigue specimens under cyclic loading conditions.Through a series of processing processes,the relationship between the change of sample entropy and the degree of damage was found.(5)The BP neural network damage phase pattern recognition method can be used to identify the damage phase pattern under constant load and cyclic load,and the wavelet soft threshold noise reduction processing data and unnoise reduction data are compared.From the curve of pattern recognition accuracy and the curve of loss function,it can be seen that the noise reduction method in this paper has higher accuracy in the identification of damage stage.