Topological optimization of rotorcraft structures for crashworthiness requirements

The optimal design of rotorcraft structures is studied for improved crash performance. While the design approach is applicable to both primary and system support structures such as landing gears and crew seats, the focus of this study resides in the design of the floor and subfloor structures with a special emphasis on topological design. A formal design methodology based on a genetic algorithm optimization approach was successfully implemented for the topological and sizing optimizations of rotorcraft floor/subfloor structures for crashworthiness considerations. In this context, the design of energy absorbing (EA) structures was formulated as determining load-deflection curves for these structures that would yield an optimal response during crash. The simultaneous sizing and placement (topology) of these structural EA components under a rigid floor structure was successfully demonstrated in a number of test problems. Practical applications of this approach are made in the crashworthy design of a generic floor section structure, and an idealized model of a CH-47A helicopter. A major contribution of this research is the use of neural network based approximate models to replace the KRASH (hybrid crash analysis code) and ABAQUS (Nonlinear FEM) models in calculating crash response with a significant attendant savings in CPU. An inverse problem solution capability was developed to recover physical dimensions of an energy absorbing concept consisting sine-wave corrugation in the web of subfloor beams, from a load-deflection curve obtained as a solution to the optimization problem.