基于希尔伯特-黄变换测量搅拌釜临界分散转速

Measurement of critical dispersion speed in stirred tank based on Hilbert-Huang transform

基于希尔伯特-黄变换(HHT)分析搅拌釜壁产生的声发射(AE)信号,获得了代表气液体系运动的特征频段(10～80kHz),将其对AE信号进行重构.以重构信号能量为特征参数,根据其随搅拌转速的规律性变化,提出搅拌釜临界分散转速的测量判据,即重构信号能量由平稳突然增大时所对应的搅拌转速为临界分散转速.与目测法相比,该方法平均相对偏差小于2.86%,具有较高的精度.研究发现,外循环破坏了搅拌釜内的径向流动,不利于气液分散,导致临界分散转速变大.根据不同外循环条件下的实验结果,建立外循环搅拌釜临界分散转速的预测关联式,平均相对偏差小于0.65%.该研究结果表明,声发射结合HHT的方法可以准确实现临界分散转速的测量.

Hilbert-Huang transform （HHT） was used to analyze the acoustic emission （AE） signals re- ceived from the wall of a stirred tank. The characteristic frequency scale （10-80 kHz）, which represented the gas and liquid movement, was obtained to reconstruct the AE signals. The energy of the reconstructed AE signals was found changing regularly with increasing impeller speed in the stirred tank. Then, a crite- rion to determine the critical dispersion speed was proposed. When the energy of the reconstructed signals begins to increase remarkably after remaining constant, the corresponding impeller speed is the critical dis- persion speed. Compared with visual observation, the average relative deviation of this method was less than 2.86%. Furthermore, it was found that external loop could fairly increase the critical dispersion speed by reducing the radial flow in the tank, and impairing the gas dispersion in the tank. According to the experimental results of different external circulation conditions, a predictive model for calculating the critical dispersion speed in a stirred tank with external loop was established. Its average relative deviation was less than 0.65%. The results show that the multi-scale analysis of acoustic signals based on HHT can accurately measure the critical dispersion speed.