Research On Monitoring Of Acoustic Emission Signal In Brittle Removal Process Based On Sparse Decomposition

At present,hard-brittle materials are more and more widely used in information,medical,semiconductor,digital technology,military,and other fields.With the increasing demand for hard brittle materials,the requirements for technical indicators such as processing accuracy and processing efficiency are also increasing.The improvement of high-quality and high-efficiency processing technology for hard and brittle materials has become a hot topic of current research.Through on-line monitoring of the processing process,the quality of processing can be effectively ensured and the processing efficiency can be improved.Acoustic Emission(AE)can reflect the formation,expansion and fracture of internal cracks in brittle materials,which is a commonly used monitoring method for processing processes.In this paper,a diamond abrasive head was used to simulate the single-grinding grinding movement,and acoustic emission technology was used to monitor the abrasion of the indenter and the scratching experiment of BK7 glass,sapphire and silicon carbide.Using the sparse-invariant sparse coding algorithm to extract features of the AE signals during the process,the time and energy information of the occurrence of cracks(such as crack generation and fragmentation)in the brittle material can be obtained,and the brittleness removal process of the material can be expressed more intuitively.Specifically carried out the following work:(1)Set up an experimental platform and signal acquisition system,analyzing the effect of AE sensor installation position on the AE signal acquisition;comparing the effect of the angle of the top of the diamond indenter on the brittle material removal effect.(2)Carry out BK7 glass slope scratching and horizontal scratching experiments.Experimental studies show that the force signal and acoustic emission signal have a linear rise relationship with the depth of scratches,the deeper the scratches,the more brittle removal phenomena such as cracks and chipping,the greater the amplitude of the fluctuation of the force signal,the higher the frequency and energy of burst-type acoustic emission signals.(3)Diamond head life-cycle experiments were carried out.Through the statistical analysis of the AE signal during the experiment and the verification of the wear profile of the indenter,the critical conditions for the tool to be trimmed or replaced were evaluated.(4)Study the theory of shift-invariant sparse coding algorithm,extract features of AE signals in the brittleness removal stage of BK7 glass,sapphire,and silicon carbide.The processing process monitoring method based on the AE feature event as the monitoring unit is proposed,and the time and size information of the brittle removal events such as material cracks and breakage can be accurately obtained.