MODELLING THE MELTING TEMPERATURE OF A LIPID‐BASED CRITICAL TEMPERATURE INDICATOR: A COMPARISION BETWEEN SIMPLEX-LATTICE AND SIMPLEX-CENTROID DESIGNS

Critical temperature indicators (CTI) find applications in food industry in cases when defrost may not occur or a specific temperature may not be reached, , indicating changes through visual changes, such as melting, color changes, etc. Lipid mixtures are promising candidates to formulate CTI since the final melting point of the mixture may be manipulated by the proportion of each lipid. In this work a lipid mixture consisting of stearic acid, lard and peanut oil was used to develop a CTI mixture. Simplex-lattice and Simplex-centroid experimental designs were compared to modelling the melting temperature of the lipid mixture. Addition of axial points to the experimental design improved predictive ability of the models while the inclusion of inverse terms was necessary to improve models accuracy. Simplex-lattice design presented an improved ability to predict the melting point of binary mixtures, while the simplex-centroid design resulted in an improved model for predicting melting point of the ternary mixtures