Propeller aerodynamic optimisation to minimise energy consumption for electric fixed-wing aircraft

Abstract

An electric propulsion model for propeller-driven aircraft is developed with the aim of minimising the power consumption for a given airspeed and thrust. Blade Element Momentum Theory (BEMT) is employed for propeller performance predictions fed with aerodynamic aerofoil data obtained from a proposed combined Computational Fluid Dynamics (CFD)–Montgomerie method, which is also validated. The Two-Dimensional (2D) aerofoil data are corrected to consider compressibility, three-dimensional, viscous and Reynolds-number effects. The BEMT model showed adequate fitting with experimental data from the University of Illinois Urbana Champaign (UIUC) database. Additionally, Goldstein optimisation via vortex theory is employed to design pitch and chord distributions minimising the induced losses of the propeller. Particle swarm optimisation is employed to find the optimal value for a wide range of geometrical and operational parameters considering some constraints. The optimisation algorithm is validated through a study case where an existing optimisation problem is approached, leading to very similar results. Some trends and insights are obtained from the study case and discussed regarding the design of an optimal propulsion system. Finally, CFD simulations of the study case are carried out, showing a slight relative error of BEMT.

Aircraft Propeller Design through Constrained Aero-Structural Particle Swarm Optimization

“…Additionally, the distributions are compared against Goldstein distributions, which are the optimal pitch and chord calculated through the Euler-Lagrange equations applied to vortex theory. Some details about how to calculate these distributions can be found in [9]. Nevertheless, the Goldstein distributions are unique for given trust, airfoil distribution, and operational conditions having a considerably lower computational cost; hence, these distributions are analyzed for looking at an insight, which allows reducing the current computational cost of searching a distribution with a PSO that fulfills the target thrust while minimizing the torque.…”

“…Nevertheless, the Goldstein distributions are unique for given trust, airfoil distribution, and operational conditions having a considerably lower computational cost; hence, these distributions are analyzed for looking at an insight, which allows reducing the current computational cost of searching a distribution with a PSO that fulfills the target thrust while minimizing the torque. Moreover, in [9], the thin propeller results also showed the difficulty of optimizing the propeller aerodynamic without considering the structural constraints. However, similar airfoil thickness and camber distributions (with a minimum value somewhere between tip and root) were obtained in that study (nevertheless, the airfoil design space is different).…”

“…For the optimization method, a Particle Swarm Optimization (PSO) routine is developed, which guarantees the fulfilment of the target performance and the constraints, including the allowable stress where a structural model based on the Euler-Bernoulli beam theory is used to evaluate the structural viability of the candidate propellers. The PSO methodology has been employed for propeller optimization in different researches [6][7][8][9], and it also has been reported to be faster than other algorithms such as Generic Algorithms or sine-cosine optimizers [10,11].…”

“…This research seeks to expand and contribute in different ways, such as the presentation of a BEMT algorithm with improved convergence and recursive routines that have not been exposed before for propellers [16], the implementation of a constrained PSO routine for a multi-disciplinary propeller optimization, the investigation of the airfoil geometry distribution along an optimal blade considering structural restrictions, proposal of the specific coefficient values of the rotational correction for small propellers (values in literature are for much larger rotors and for wind turbines that operate at considerably lower rotational speed [17][18][19]), testing of the feasibility of printed propellers, proposal of an equation to correct the stalling angle for different Reynolds numbers, validation through FEA the Euler-Bernoulli beam theory applied for propellers, and conclusion about the suitability for a BEMT fed by XFOIL and CFD. Moreover, some previous works do not manufacture the optimal solutions to validate the results or do not consider the interaction of the propeller efficiency with the motor efficiency [2,9].…”