Predicting the velocity distribution of Rushton turbine impeller in mixing of polymeric liquids using fuzzy neural network models

Abstract

Velocity profiles are helpful for the confident design of mixing tanks and chemical reactors in mixing processes. A fuzzy model and an artificial neural network have been presented for accurate prediction of velocity distribution of Rushton turbine impeller (RTI) for the mixing of polymeric liquids in the lower transition region: 35<Re′<1800. Local tangential and radial velocities were predicted along the discharge plane of the impeller. Experimental data were used for training, validation, and testing the neuromorphic models. The presented models are very accurate and reliable in predicting the velocity profiles over wide ranges of polymer concentrations and rotational speed. Comparison of the suggested fuzzy model and the empirical correlations shows that the proposed model outperforms the other alternatives both in accuracy and generality. The results show that the proposed neuromorphic models can successfully be used for prediction of velocity distribution in agitated tanks for viscoelastic polymeric fluids.