Modeling and optimization of pyrotechnic compositions using central composite design and response surface methodology

This study focuses on optimizing pyrotechnic compositions containing boron (B) and boron carbide (B 4 C) as fuels and bismuth oxide (Bi 2 O 3 ) as the oxidizer. A Central Composite Design (CCD) and Response Surface Methodology (RSM) were employed to screen and evaluate the effects of component concentrations on combustion parameters, specifically adiabatic temperature and constant-volume pressure. Simulations using REAL WIN provided the necessary data, which were analyzed to derive predictive models. Optimization was conducted for various purposes, including matching specific targets, maximizing or minimizing outputs, and balancing performance criteria, demonstrating the flexibility and utility of the approach. Results obtained for the optimal compositions showed excellent agreement with the simulation, with prediction errors below 2% in most cases. For instance, an optimized composition (9.53% B, 3.91% B 4 C, 86.56% Bi 2 O 3 ) matched a target temperature of 2000 K with an error between RSM prediction and REAL WIN simulation lower than 1.4% for the constant-volume pressure. The study further explores the limitations and opportunities for improving the model by expanding the experimental range and incorporating additional data points. This investigation highlights the potential for refining predictive accuracy while addressing challenges associated with nonlinear behaviors and extrapolation. By combining DoE approaches with simulation tools, this work offers reliable guidance for formulation and experimental planning in energetic materials development.