Studies on Batch Drying Characteristics and Transfer Coefficients of the Food Grains and Vegetables in A Fluidized Bed Dryer using Ann Method

Abstract

Attempts have been made to study the drying characteristics for different grains and vegetables by carrying out experiments in a fluidized bed dryer. Drying characteristics have been analysed in terms of moisture content and diffusivity for grains and vegetables dried through the laboratory scale fluidized bed dryer. Characteristic drying curves have been plotted. The effects of different system parameters (viz. temperature, time, and density of material and fluid velocity) on the drying performance i.e. on moisture content and diffusivity of the sample have been studied in the present work. Temperature dependence of mass transfer coefficients have been established in terms of Arrhenius type of relation. Drying curves of different samples indicate that drying occurs mostly in falling rate period due to internal diffusion as effective diffusivity of different samples were found to decrease with increase in drying time. Fick’s second law of diffusion is found to be applicable for the drying process during the falling rate period. Attempt has also been made to develop correlations for the moisture content and diffusivity of the samples relating the respective experimentally observed data with the different system parameters on the basis of dimensional analysis which have been validated by Artificial Neural Network analysis. Finally the calculated values of the moisture content obtained through the Dimensional analysis and Artificial Neural Network analysis have been compared with the experimentally measured values. Standard deviations and mean deviations for different samples were found to be less by both Dimensional analysis and ANNanalysis respectively. Very good agreement between the calculated values by ANN-analysis and experimental data proves that the Neural Network training is proper and it has learned well the behaviour of the different parameters.