Abstract
In this paper, acoustic emission data fusion based on multiple measurements is presented for damage detection and identification in oxide-based ceramic matrix composites. Multi-AE (acoustic emission) sensor fusion is considered with the aim of a better identification of damage mechanisms. In this context, tensile tests were conducted on ceramic matrix composites, fabricated with 3M™ Nextel™ 610 fibers and aluminosilicate matrix, with two kinds of AE sensors. Redundant and complementary sensor data were merged to enhance AE system capability and reliability. Data fusion led to consistent signal clustering with an unsupervised procedure. A correlation between these clusters and the damage mechanisms was established thanks to in situ observations. The complementarity of the information from both sensors greatly improves the characterization of sources for their classification. Moreover, this complementarity allows features to be perceived more precisely than using only the information from one kind of sensor.