Determination of volumetric mass transfer coefficient in gas-solid-liquid stirred vessels handling high solids concentrations: Experiment and modeling
Document Type
Article
Publication Date
1-1-2018
Abstract
Rigorous analysis of the determinants of volumetric mass transfer coefficient (kLa) and its accurate forecasting are of vital importance for effectively designing and operating stirred reactors. Majority of the available literature is limited to systems with low solids concentration, while there has always been a need to investigate the gas-liquid hydrodynamics in tanks handling high solid loadings. Several models have been proposed for predicting kLa values, but the application of neuro-fuzzy logic for modeling kLa based on combined operational and geometrical conditions is still unexplored. In this paper, an ANFIS (adaptive neuro-fuzzy inference system) model was designed to map three operational parameters (agitation speed (RPS), solid concentration, superficial gas velocity (cm/s)) and one geometrical parameter (number of curved blades) as input data, to kLa as output data. Excellent performance of ANFIS’s model in predicting kLa values was demonstrated by various performance indicators with a correlation coefficient of 0.9941.
Keywords
Adaptive neuro-fuzzy inference system, Artificial intelligence-based modeling, Artificial neural networks, Stirred vessels, Volumetric mass transfer coefficient
Divisions
fac_eng
Publication Title
Iranian Journal of Chemistry and Chemical Engineering
Volume
37
Issue
3
Publisher
Iranian Research Institute of Development in Chemical Industries