|
Abstract
|
Based on 3-D elasticity approach, differential quadrature method (DQM) in axial
direction is adopted along with Globalized Nelder–Mead (GNM) algorithm to
optimize the stacking sequence of a laminated cylindrical shell. The anisotropic
cylindrical shell has finite length with simply supported boundary conditions.
The elasticity approach, combining the state space method and DQM is used to
obtain a relatively accurate objective function. Shell thickness is fixed and
orientations of layers change in a set of angles. The partial differential equations
are reduced to ordinary differential equations with variable coefficients by
applying DQM to the equations, then, the equations with variables at discrete
points are obtained. Natural frequencies are attained by solving the Eigenfrequency
equation, which appears by incorporating boundary conditions into the
state equation. A GNM algorithm is devised for optimizing composite
lamination. This algorithm is implemented for maximizing the lowest natural
frequency of cylindrical shell. The results are presented for stacking sequence
optimization of two to five-layered cylindrical shells. Accuracy and convergence
of developed formulation is verified by comparing the natural frequencies with
the results obtained in the literature. Finally, the effects of mid-radius to
thickness ratio, length to mid-radius ratio and number of layers on vibration
behavior of optimized shell are investigated. Results are compared with those of
Genetic Algorithm (GA) method, showing faster and more accurate convergence
|