Application Research Of Acoustic Emission Technology In Crumb Rubber Concrete

Acoustic emission technology is a commonly used real-time monitoring technology.It has been widely used in many fields such as petrochemical industry,aerospace industry,scientific research and material testing industry.Research in the civil engineering field is mostly used for concrete and rock materials.Because acoustic emission technology has unique advantages in the research of concrete materials,and crumb rubber concrete materials have great development prospects in the future,so the damage process and damage mode of crumb rubber concrete materials by acoustic emission technology are used.Research is of great significance for the application and detection of such new concrete materials in practical engineering in the future.In view of this,this paper use acoustic emission technology to study rubber aggregate concrete.The main research contents include:Starting from the function of acoustic emission technology for real-time monitoring,the characteristic parameters of the cumulative hits number of the four-point bending experimental acoustic emission data of crumb rubber concrete beam are collected,and the micro-crack monitoring is carried out by acoustic emission technology.The crumb rubber concrete failure process is divided into three stages by the trend of cumulative number of hits.At the same time,the acoustic emission data are analyzed by Ib-value method and acoustic emission signal strength analysis method using real-time numerical values to characterize the severity of the damage process..The crack failure mode of crumb rubber concrete are characterized and analysed using k-means cluster analysis method and RA-AF value analysis method to distinguish between the tensile crack and the shear crack.Starting from the function of acoustic emission technology for damage pattern recognition,this paper apply and improve a new clustering method--fast search and find of density peaks clustering method.This method is compared with the traditional four clustering methods(fuzzy c-means clustering method,GMM clustering method,SOM clustering method and hierarchical model clustering method).The best clustering method suitable for crumb rubber concrete is determined.Combined with the results of cluster analysis,the damage pattern recognition of crumb rubber concrete is carried out by means of fast Fourier transform and peak frequency analysis to distinguish between different damage failure modes(concrete matrix rupture and rubber fracture).Starting from the function of acoustic emission technology for being capable of predicting crack propagation,based on the acoustic emission parameters,the artificial neural network is used to predict the ultimate load of the crumb rubber concrete beam under four-point bending condition.An acoustic emission multi-parameter artificial neural network prediction model is proposed.The maximum prediction error of this model is only 0.89%.