Modular Design and Optimization Framework for Pressure-Fed Rocket Upper Stages

This work presents a modular framework for the design and optimization of pressure-fed engines for liquid-propellant upper stages. Unlike conventional methods, an innovative subalgorithm for the optimization of the inboard profile is implemented. This tool aims to minimize the inert stage mass by identifying the optimal tank configuration-shape, arrangement, and number-that satisfies specific geometric constraints. The framework combines analytical models for the main propulsion and stage subsystems with empirical mass estimation relations to improve the accuracy of mass predictions. A multi-objective optimization strategy, based on a genetic algorithm, minimizes both dry and propellant masses by varying chamber pressure, mixture ratio, and expansion ratio, while respecting mission-derived requirements such as the target Delta V. The analysis of the resulting Pareto fronts highlights clear tradeoffs between propellant mass, structural mass, and engine performance, as well as the influence of geometric parameters like stage radius and length-to-radius ratio on the optimal configurations. Comparison with 10 engines and 6 upper-stage datasets resulted in maximum deviations of 2.7% in dry mass and 3.6% in total mass compared to reference data, suggesting that the tool can reproduce real configurations with very good accuracy.