Surrogate-model based Optimization Framework for the Structural Design of Helicopter Rotor Blades subject to Dynamics and Failure Criteria

Abstract

Helicopter rotor hub vibratory loads can be alleviated through careful design of the rotor blade inner structure. Large design space and the nonlinear nature of the problem are major obstacles that need to be overcome. Apart from that, the need for high-fidelity solutions lead to high computational times. An automated surrogate-model based design optimization process using commercially available software as well as codes developed within DLR has been described in this paper. Minimal human interference and overall process efficiency is the goal of this work. Latin Hypercube Sampling function for the design of experiments, Kriging function for surrogate modeling and particle swarm optimization algorithm make up the framework. The rotor blade inner structure parameters constitute the design variables while the natural frequencies and vibration index are taken as objective functions. Results show that through careful design of the inner structure, it is possible to obtain a rotor with lower hub vibratory loads. For the optimization process, it must be ensured that sufficient number of sampling points are taken for building accurate surrogate models and the problem definition should be neither under-constrained nor over-constrained.