Characteristic Extraction Of Acoustic Emission Signals Of Refractory Materials During Damage Test

Refractory materials are the indispensability components of high temperature vessels, andthe key to ensure the safe operation of high temperature equipment. The microscopic damageform and degree of materials determine their macroscopic properties. Microscopic damageexperiment characterization has a great significance for studying microstructure influence on theproperties of materials. The acoustic emission technology (AE) was used to investigate thedamage mechanism of MgO-C refractory under three-point flexural tests. AE characteristicparameters and waveforms were analyzed using parameters analytical method, wavelet transformand independent component analysis. Features of AE signals were obtained. Moreover, themicro-mechanism and damage form during the loading damage process were researched. Themain work and conclusions of this dissertation are as follows:(1) The three-point flexural tests of MgO-C refractory was performed using differential hightemperature stress and strain machine. AE characteristic parameters and waveforms werecollected by acoustic emission testing equipments.(2) The AE energy and centroid frequency of power spectrum time-history chart of thesignals were analyzed using parameters analytical method. Damage evolution process of MgO-Crefractory includes four main stages: initial damage period, damage evolution period, accelerateddamage period and specimen overall fracture period. The damage mechanism and characteristicsin each damage phase of MgO-C refractory materials were derived in the loading damageprocess.(3) AE signals of MgO-C refractory during different damage stages have been analyzed bywavelet transform combined with independent component analysis, and actualize the blindsource separation and feature extraction of different damage source signals. The results show thatdamage patterns of MgO-C refractory includes two main types: matrix damage and interfacedamage. The characteristic frequencies of matrix damage is7.8~15.6kHz (matrix crack) and32.5~62.5kHz (matrix crack propagation); the characteristic frequencies of interface damage is62.5~250kHz (interface debonding). Wavelet transform combined with independent componentanalysis for blind source separation and feature extraction of AE signals is effectively. Theexperiment characterization of microscopic damage of refractory materials based on acousticemission detection has certain reference value for the nondestructive testing of large industrialfurnace, and laid a foundation for further study on the theory relation model of microscopicstructure and macro properties of refractory materials.