GEOSTATISTICAL MODELS USED IN GRAIN HYDRATION

Résumé

Empirical models that present a behavior of stabilization (reaching a constant level) are widely used in Geostatistics (a branch of statistics that deals with spatial variability). These models can be applied to data which varies in time and tends to stabilize. Therefore, the exponential model and the powered exponential model were applied in the case of grain hydration kinetics. Both models were fitted to experimental data on moisture content as a function of time at five different hydration temperatures, using the least squares method to obtain their parameters. The times required for the grains to reach moisture content close to equilibrium were obtained using a method which corroborates the results when the model reaches 99.99% of the moisture content constant level. The results have shown that the powered exponential model is the most adequate one for describing the main trends of the experimental data. Equilibrium times obtained in the present work were much closer to experimental data than those obtained using the traditional approach, which considers the curve to be stable at 95% of its final constant level.