Multi-objective industrial optimization of high-speed helicopter main rotor blades with dynamically-adapted structural properties

Thumbnail Image
Desvigne, D.
Coisnon, R.
Michel, B.R.
Thomas, A.
Pinacho, J.P.
Roca León, E.
Journal Title
Journal ISSN
Volume Title
A multidisciplinary multi-objective optimization tool chain developed for designing main-rotor blades of high-speed helicopters is presented. Objectives focus on the dynamic loads in high-speed level flight, along with rotor power in hover. The tool chain relies on a genetic algorithm coupled to loads and performance simulation tools. In order to fulfill potentially strong industrial constraints, low-cost models are used in the comprehensive rotor simulation code HOST. Dynamic control loads may become critical in high speed and accurately representing both inertial and elastic responses of the blade to aerodynamic excitations is essential in the blade-design process. A new methodology is proposed in this work for the modelling of the blade structural properties. The strategy adopted is as follows: the design space is first sampled with a fixed number of donor-blade designs which cover most of the foreseen realistic planforms. The internal structure of each donor blade is tailored so that the blade eigenfrequencies placement is optimal with respect to the rotor harmonics, together with a mitigation of the blade-to-cabin modal transmissibility. Then, for each design candidate in the optimization process, a relevant donor blade is selected amongst the donor pool based on geometric similarities. The candidate structural properties are finally obtained by correcting the donor structural description with empirical functions based on the geometric discrepancies, assuming a similar structural technology. Twist, chord, and offset spanwise distributions are parameterized with Bézier curves and an optimization procedure is conducted so as to minimize the required power for hovering as well as to minimize the dynamic control loads in highspeed cruise conditions. A Pareto optimal solution is chosen and a more detailed analysis is achieved using higher-fidelity tools. The inclusion of the automatic structural update provides more realistic blade designs. The analysis of the updated structural properties indicates that the hypothesis assuming that the donor and the candidate blade have very similar eigenmodes is not entirely verified. Nevertheless, the method yields promising candidates and demonstrates the challenge of integrating the blade internal properties into the global industrial design process.