PREDICTION OF MASS TRANSFER COEFFICIENT IN BUBBLE COLUMN USING ARTIFICIAL NEURAL NETWORK

Abstract

The volumetric mass transfer coefficient kL.a was calculated using two gases (air and CO2) in water and NaOH solution. The experiments were carried out using 0.1 m column diameter. Empirical and Artificial Neural Network (ANN) correlation were developed to predicted mass transfer coefficient in form of dimensionless groups (Sh, Re,Bo and We). The use of Back Propagation Neural Network (BPNN) gave better results than other correlations found in literature and than the empirical one found in this study.