Modeling of drying rate of non-hygroscopic products: Application to the pozzolana drying

Polynomial model adopted for the moisture content influence to the drying rate of non-hygroscopic products like the pozzolana presented some limits. This model was not universal and has diverged whether moisture content different to the appropriate experiment values has been used. Thus, a new exponential model based on the inverse of moisture content and respecting to the imposed boundary conditions has been proposed. Model parameters were determined using the least square approximation method of Newton-Marquart. The predicted temporal variations of moisture content were provided from the Runge-Kutta fourth order integration scheme computed on the new model of drying rate. Moreover, two-way ANOVA method was performed on data analysis, at the parameters statistical significance and in residual analysis for the model fit quality. Model validations were done by comparisons among theoretical and experimental values and between predicted and polynomial models. Comparisons showed a good agreement with highest R2 and lowest reduced chi-square, MBE, RMSE, and MAPE.