Chemical Reactor Network for Hybrid Rocket Engine Optimization

This study proposes the integration of a chemical reactor network into the design optimization procedure of hybrid rocket engines. The adoption of a chemical reactor network enables a more realistic representation of combustion phenomena compared to conventional equilibrium-based formulations by embedding nonideal effects directly into engine design and performance evaluation. The combustion chamber is discretized into four perfectly stirred reactors (oxidizer-core, fuel-rich, flame, and mixer), implemented within the open-source framework Cantera. The network is trained using a particle swarm optimization algorithm against experimental data from the literature on gaseous oxygen and paraffin-based wax propellants, demonstrating high accuracy and predictive capability. The trained chemical reactor network model provides a means to compute the actual characteristic velocity efficiency for each engine configuration considered during optimization, thus improving performance and mass computation during ascent integration. The proposed approach is applied to a reference hybrid rocket upper-stage mission, comparing the performance with and without mixing-enhancing devices. Results indicate that a framework based on a chemical reactor network offers a robust foundation for coupled engine–trajectory optimization, enhancing the physical consistency and reliability of hybrid rocket propulsion system design.