Damage pattern recognition of Refractory Materials based on k-means clustering analysis

摘要

The k-means algorithm was used to divide the acoustic emission signals collected during the three-point bending test into two types. Combining with the analysis of AE parameters can we distinguish the micro-damage pattern recognition of the refractory materials. The bending test equipment is HMOR/STRAIN, and the AE acquisition device is DISP from PAC. Amplitude, counts, rise-time, duration and centroid frequency were selected as the AE parameters. The microscopic damage modes of the refractory materials were recognized.