A Multidisciplinary Design, Analysis, and Optimization (MDAO) Framework for Conceptual UAV Design

This study presents a Multidisciplinary Design, Analysis and Optimization framework for the design of electric unmanned aerial vehicles (UAVs) at the conceptual design phase. It is applicable to various medium class UAV configurations, including conventional, tandem, cruciform, delta wing etc. The framework integrates multiple design and analysis tools within a single workflow. An airfoil design tool (XFOIL), a CAD design tool (OpenVSP), an aerodynamic analysis solver (FlightStream, a surface-vorticity-based panel code), and an optimization toolbox (OpenMDAO). It incorporates a direct weight estimation model to reduce divergence issues commonly encountered with empirical formulations. A sensitivity analysis is first performed using grid sampling to determine the effects of design variables on the flight performance, weights, stability, while ensuring structural integrity under maximum loading flight conditions. This reduces the number of design variables for subsequent optimization by removing the parameter which are least influential and also minimizing gradient noise. The optimized UAV design is then obtained using a gradient-descent method. The final design is validated through a Reynolds-averaged Navier–Stokes (RANS) simulation in ANSYS Fluent. The comparison shows strong agreement in the linear region of the flight regime. This proposed framework provides reliable and consistent design choices to the designer, producing configurations comparable to existing UAVs in the market. It reduces the time required in the conceptual design phase while reducing reliance on empirical correlation and thereby providing a physics-based foundation at early-stage UAV development.