|
We obtained the characteristic scale (G1) of the acoustic emission (AE) signals representing the correlations between bubbles and liquid phase in a gas-liquid stirred tank by using multi-scale wavelet transform and R/S analysis. According to the observation that the energy fraction of AE signals in G1 scale changed regularly with the increasing impeller speeds, a method to determine the flooding/loading transition was presented. The critical speed that represents the regime transition, Nf, was associated with the emergence of steady state in AE energy fractions in G1 scale. Through experiments taken out under different conditions (flow rate, static liquid height and impeller fixed height), AE method based on multi-scale analysis was proved to show adequate sensibility and feasibility in the range of experimental conditions analyzed. By comparing experimental results from AE measurement with that from visual and power draw methods, we found that the average relative errors of AE method were no more than 2.62 % and 1.38%, respectively. It is concluded that multi-scale analysis of AE signals is a feasible means for determining the flooding/loading transition and have the merit of being sensitive, non-intrusive and accurate. |
|
Keywords:Acoustic emission; Multi-scale; R/S analysis; Flooding/loading; Flow regime |
|