Aerodynamic Optimization of a Low-Altitude Airship With Distributed Electroaerodynamic Propulsion

This study presents an automated Bayesian Optimization framework for the aerodynamic layout optimization of distributed electroaerodynamic (EAD) thrusters on a low-altitude airship envelope. The framework couples a RANS solver with a Gaussian Process surrogate and an Expected Improvement acquisition function, enabling systematic exploration of a four- dimensional design space, comprising longitudinal and angular thruster positions, with a limited number of high-fidelity CFD evaluations. The optimization identifies the longitudinal positions as the dominant design variables, with performance differences across configurations traced to the pressure drag on the outer nacelle wall. The optimal configuration is located approximately at x/L = 0.26, where x is the longitudinal coordinate measured from the nose and L is the airship length, yielding a 61% improvement in the net force coefficient over the best configuration sampled during the initial Design of Experiment. This optimum arises from the interplay between nacelle pressure drag and the streamwise projection of the thrust vector, producing a well-defined interior optimum within the axial design domain. The framework converges within 100 CFD evaluations beyond the initial Design of Experiments, with the surrogate achieving a coefficient of determination greater than 0.99 on the training set.