Ternary Blends of Vegetable Oils: Thermal Profile Predictions for Product Design

This work deals with Product Design by means of theoretical predictions of the Solid Fat Content of different formulations using 3 vegetable oils. A Soli-Liquid Equilibrium (SLE) model was implemented and integrated into an optimization algorithm based on the Generalized Reduced Gradient method. A total of 3,696 SLE problems are solved, covering 57 binary blends, 3 pure vegetable oils and 171 ternary blends problems, before and after chemical interesterification reaction and at 8 different temperatures. A combinatorial random distribution of fatty acids in the glycerol structure is used to simulate the effect of the reaction. The results were compared with 256 experimental points, giving an average absolute error of 5.4 and 4.4 in Solid Fat Content for systems before and after reaction, respectively. Computer-aided tools can be useful to deal with the large combinatorial problem faced by product design, especially when desired product performance is related to a phase behavior in multicomponent mixtures.