基于超声的高浓度浆液两相流粒径在线测量

Online measurement of particle size of high concentration slurry two-phase flows based on ultrasound method

近年来颗粒粒径测量问题受到了国内外的广泛研究,其中高浓度两相流的在线测量越来越成为关注的焦点。为测量石灰石浆液在流动过程中的颗粒粒径分布,由超声在线测量装置获得质量分数10%~40%浆液在不同时间的超声衰减谱。根据Harker&Temple和BLBL散射叠加模型预测,结合基于最优正则化的加权非负最小二乘算法(ORT-WTLS)计算浆液粒径分布。结果表明:由不同粒径理论衰减谱和10dB含噪衰减谱计算的颗粒直径与设定值的相对误差均小于5%。衰减系数并非随浆液浓度线性递增,表明Harker&Temple和BLBL非线性模型适用于高浓度浆液体系的理论预测。使用ORT-WTLS算法对浆液实验衰减谱中的测量误差作加权处理并计算颗粒粒径分布,其测量结果与图像法基本吻合,动态测量误差小于7%。因此结合实验超声衰减谱、颗粒测量的非线性声学理论模型及ORT-WTLS反演算法可准确计算高浓度浆液的颗粒粒径分布。

In recent years,particle size measurement has been widely studied at home and abroad,and the online measurement of high concentration two-phase flow of slurry has become the focus of attention.In order to measure the particle size distribution of limestone slurry during flow,ultrasonic attenuation spectra of slurry with mass concentrations of 10%—40%at different moments were obtained by an ultrasonic online measuring device.A weighted non-negative least squares algorithm(ORT-WTLS)based on optimal regularization was used to calculate the slurry particle size distribution based on the predicted theoretical spectra of Harker&Temple and BLBL scattering models.The results showed that the relative deviations of the particle diameters calculated from the theoretical attenuation spectra and the 10dB noise-containing attenuation spectra of different particle sizes were less than 5%.The attenuation coefficients did not increase linearly with slurry concentration,indicating that the Harker&Temple and BLBL nonlinear models were suitable for the theoretical prediction of high concentration slurry systems.The ORT-WTLS algorithm was used to weight the measurement errors in the slurry experimental attenuation spectrum and calculate the particle size distribution,and the measurement results were basically consistent with the image method,and the dynamic measurement deviation was less than 7%.Therefore,combining the experimental ultrasonic attenuation spectrum,the nonlinear acoustic theoretical model of particle measurement,and the ORT-WTLS inversion algorithm can accurately calculate the particle size distribution of highly concentrated slurry.