Development of Software for Predicting the Thermophysical Properties of Foods Based on Their Composition Using the Choi–Okos Model.

Abstract

The thermophysical properties of food vary with temperature and composition; therefore, estimating them is essential for modeling heating, cooling, refrigeration, and freezing processes. These elements enable this study, which aims to develop software to estimate the specific heat (C_p), density (rho), thermal conductivity (k), and thermal diffusivity (alpha) of food based on its proximate composition and temperature. This is achieved through the computational implementation of correlations based on the Choi-Okos model. The application was developed in Python using Flask and SQL Alchemy, with moisture, carbohydrates, proteins, fats, fiber, ash, and temperature as input variables. It accepts temperatures between -40 °C and 150 °C, restricting each component to the range of 0 to 100% and verifying that the total composition sums to 100% before running the calculations. The computational core calculates the individual properties of the main constituents as a function of temperature and automatically generates the food's thermophysical values. It also stores, displays, and exports the results. Functional testing confirmed the integrated operation of the recording, processing, query, and export modules. However, a mathematical review of the algorithm revealed the need to refine certain mixing rules and the handling of frozen water before establishing the model's final predictive accuracy. The developed software constitutes a functional computational platform for composition-based thermophysical estimation.