Influence of Cultivar on the Predictive Performance of a Moisture Transport Model Developed for Parboiled Paddy Drying

Abstract

A mathematical model for predicting moisture profile in parboiled paddy during drying was developed using the prolate spheroidal coordinate system. The model was solved using the Crank Nicolson central difference scheme and the variation of its predictive potential with paddy cultivar was investigated. Parboiled paddy drying was carried out at 60°C using a cabinet tray dryer. The variation of the moisture ratio of different paddy cultivars (Banki, De-gold, and Liberia) with time and drying rate and with time and moisture content was determined. Model validation was carried out using residual plots, root mean square deviation, and coefficient of determination. The parboiled Banki, De-Gold, and Liberia paddy cultivars had equivalent diameters of 3.73, 4.75, and 3.68 mm, prolate spheroidal geometric dimensions (equatorial and polar radii) of 1.16 and 4.78 mm, 1.22 and 5.45 mm, and 3.10 and 3.92 mm, and effective moisture diffusion coefficient of 2.81934 × 10−10, 5.71515 × 10−10, and 3.77336 × 10−10 m2/s, respectively. Drying of parboiled paddy occurred in the falling rate phase with initial period of induction and critical moisture content that varied with cultivar. The residual plots showed that the model was acceptable for predicting moisture transport in all the cultivars during drying. Comparison of the observed and predicted moisture ratios showed low root mean square deviation and high coefficient of determination, which varied with cultivar indicating that the predictive potential of the model was cultivar dependent.