DIFFUSION STUDY OF NaCl AND KCl AT THE SOLUTION/OLIVE INTERFACE: MATHEMATICAL MODELING USING THE FINITE ELEMENT METHOD AND SELF-ORGANIZING FEATURE MAP (SOFM)-TYPE NEURAL NETWORKS

Olive samples were subjected to the salting process in brine containing 1196 mol m⁻³ of NaCl and 402 mol m⁻³ of KCl. Samples were collected during 60 h and salt concentration values were determined. With the finite element method (FEM) and the minimization of the errors percentage between the simulated and experimental values, the mass transfer coefficients in the film (h_m) were obtained, being 1.0072 × 10⁻⁸ and 1.2499 × 10⁻⁸ for NaCl and KCl, respectively. The salts concentration at the olive/brine interface was simulated by FEM and analyzed via SOFM. A network with 4 × 4 topology and 10000 training epochs was used. It was observed that the influence of the stationary film formed on the surface of the olives during the salting process depends on the position, the salt involved in the diffusion and that the concentration of the salts, at each point, varies according to the immersion time.

Keywords:
self-organizing maps; NaCl; KCl; simulation; salt diffusion