Structural Optimization of Laminated Composite Blade Using Particle Swarm Optimization
Composite blades working underwater experience complicated loading conditions. Robust design of a composite blade for hydrokinetic applications should satisfy varying loading conditions and conservative failure evaluations. Blade manufacturing using composites requires extensive optimization studies in terms of composite materials, number of layers, stacking sequences, ply thickness and orientation. In the current study, particle swarm optimization (PSO) technique is adopted to conduct composite lay-up optimization for the turbine blade. Layer numbers, ply thickness and ply orientations are optimized using standard PSO (SPSO) to minimize weight. Composite failure criteria are applied using finite element method to generate the most conservative blade design. Based on the blade lay-up design with minimized weight, stacking sequence of the blade lay-up was optimized to maximum safety factor of the designed blade using permutation discrete PSO (PDPSO). To improve the efficiency of the algorithm, the concepts of valid/invalid exchange, and memory checking were introduced into PDPSO. Meanwhile, another discrete PSO using partially mapped crossover (PMX) technique was used to validate the simulation results optimized by PDPSO. A final composite blade design with minimized weight and maximized load-carrying capacity was presented.
H. Li and K. Chandrashekhara, "Structural Optimization of Laminated Composite Blade Using Particle Swarm Optimization," ASME Proceedings: Mechanics of Solids, Structures and Fluids, American Society of Mechanical Engineers (ASME), Jan 2012.
The definitive version is available at https://doi.org/10.1115/IMECE2012-88313
ASME 2012 International Mechanical Engineering Congress and Exposition
Mechanical and Aerospace Engineering
Keywords and Phrases
Composite Materials; Structural Optimization; Blades; Particle Swarm Optimization
Article - Conference proceedings
© 2012 American Society of Mechanical Engineers (ASME), All rights reserved.