Optimization of a Beverage Using Mixture Design With Operational Variables

The optimization of formulated products often requires simultaneous consideration of mixture composition and processing conditions, as both factors can strongly influence the final response. In this study, a mixture experimental design incorporating process variables was applied to the optimization of a prepared beverage (margarita), with the objective of maximizing sensory response. Four mixture components were evaluated under the combined influence of two process variables, and the experimental data were analyzed using regression modeling and analysis of variance. A stepwise regression procedure was employed to identify the most significant terms, revealing that linear mixture effects alone were insufficient to explain the system behavior. The final model exhibited excellent goodness of fit and predictive performance, with high coefficients of determination and a low residual error, while diagnostic analyses confirmed that the assumptions of normality, homoscedasticity, and independence were satisfied. Graphical tools, including simplex design plots, contour plots, and mixture surface plots, provided clear interpretation of the nonlinear response behavior and identified interior regions of the mixture space associated with favorable responses. Response optimization yielded a balanced formulation and specific process conditions that maximized the predicted sensory score with high desirability. Overall, the results demonstrate that integrating process variables into mixture design provides a robust and practical framework for beverage optimization, offering improved interpretability and decision‐making compared to traditional mixture‐only approaches.