Abstract
This paper describes a 3-stage simulation method to i/extract from experimental results the best validated empirical models of drying kinetics, ii/establish the correlations between the model coefficients and the drying airflow parameters of temperature and velocity, and the sample thickness, and then ii/use other experimental results to compare and confirm the identified model. This simulation study was applied to the case of pumpkin slices using the Response Surface Methodology (RSM) to describe the moisture ratio MR versus time. Seven thin-layer drying models including Newton, Page, Modified Page, Handerson and Pabis, Logarithmic, Midilli-Kucuk and Approximation of Diffusion models were fitted to experimental data, using nonlinear regression. It has been found that Approximation of Diffusion, Page, Midilli-Kucuk yielded the best fit. Then, best three models were selected and examined intensively, for slice thickness ranged between 0.2 and 1.4 cm, within the airflow temperature ranges (40–80 °C) and velocity (2–15 m/s). Midilli-Kucuk model gave the best correlation between the experimental and estimated data. The relationships between the model parameters (k, n, a, and b) and the drying conditions, slice thickness, and time were determined. Thus, this empirical Midilli-Kucuk thin-layer drying kinetic model including the drying conditions can accurately described with a good fitness predict and simulate the moisture ratio value for a drying process of pumpkin slices.
