Agglomerations could be detected in a fluidized bed by utilizing wavelet analysis

on the basis of the mechanism that AE signals with different frequencies were emitted when the particles with different sizes impacted the wall.The acoustic signals of particles were investigated in an externally heated fluidized bed φ100 in diameter and in a pilot-scale industrial fluidized bed φ420 in diameter respectively.In those experiments

the number of small particles decreased obviously while the number of large particles increased due to agglomeration.The signals were decomposed by multi-scale wavelet analysis.While the acoustic energy ratio increased from low scale to high scale layer by layer

the main frequency of acoustic signals transferred from high frequency to low frequency.Meanwhile

the d6th

d7th scales

which represented large size particles

increased suddenly

indicating the occurrence of agglomerations.Thus

the agglomerations in fluidized bed could be quickly and exactly detected online by the acoustic emission technique.