PropOp: Multi-disciplinary Optimization Tool for the Design Phase for a UAV Propeller Design-Fabrication-Test Stream

The increasing commercial adoption of unmanned aerial vehicles (UAVs), and in particular small UAVs weighing less than 55 pounds has led to the need of propeller designs that operate for a multitude of applications and flight... [ view full abstract ]

The increasing commercial adoption of unmanned aerial vehicles (UAVs), and in particular small UAVs weighing less than 55 pounds has led to the need of propeller designs that operate for a multitude of applications and flight regimes. Performance is dependent on the mission requirements in that the conceptual design goals drive the operating state, physical propeller geometry and ultimately determine its effectiveness. Though historically propellers are optimized to produce maximum efficiency or minimum induced losses, other performance metrics such as acoustic footprint, structural performance, manufacturability and cost are increasingly relevant and in some cases can rank higher in importance.

In this paper a multidisciplinary optimization design process for UAV propellers is presented, which considers the aerodynamic, structural and aero-acoustic design. The purpose of this is to create the design portion of a design-fabrication-test stream where propellers can be produced within a small time frame for significantly different mission requirements. The optimization program uses a simple genetic algorithm and multi-discipline feasible architecture to output propeller geometry given user prescribed flight conditions, objectives and physical design constraints.

Two solution methods are created; one where an initial solution and accompanying boundary conditions are supplied, and one which requires only input boundary conditions. Supplying an initial solution speeds up the optimization process significantly as it decreases the solution space and starts the optimization process in a local minimum. In the context of propellers this is a propeller that has a converging solution, in other words, can physically produce thrust at the prescribed RPMs, input power settings, and is continuous and homogenous. The first method is meant to be used for vehicles that have low propeller variability (i.e. strict geometrical constraints, limited engine power output or minimum required thrust). The second method is designed to find the best possible solution of the objective function given all combinations of the design variables, the global minimum, and is computationally expensive.

The results of the optimization are compared to the results of the minimum induced loss designed propeller using the revised Prandtl and Betz condition outlined in Larrabee’s method which optimizes for aerodynamics only. The benefits of this new multi-disciplined approach will be highlighted with respect to its role as a tool in the UAV propeller design-fabrication-test stream.

Preliminary results for minimizing required power at a prescribed thrust in cruising flight have yielded propeller configurations shown in Fig.1. In these cases, the number of blades, tip radius, hub radius and maximum/minimum chord lengths were specified for five blade stations and optimized for aerodynamics. Future versions plan to accommodate the aero-acoustic module and discrete variables.