OPTIMUM DESIGN OF ANTI-BUCKLING BEHAVIOUR OF THE LAMINATED COMPOSITES BY DIFFERENTIAL EVOLUTION AND SIMULATED ANNEALING ALGORITHMS

Laminate composites can be used quite naturally in automotive, marine, aviation and
other engineering branches. Determination of buckling load capacity under in-plane composite
loads of composite plates is very important for the design of composite structures. In this thesis,
Differential Evolution (DE) and Simulated Annealing (SA) are utilized for optimal stacking
sequence of a laminated composite plate, which is simply supported on four edges and is
subjected to biaxial in-plane compressive loads. The optimization problem parameters are
defined as
(i) Objective function: critical buckling load factor,
(ii) Constraints: symmetric and balanced structure, thin plate assumption, specially
orthotropic material, discrete search space,
(iii) Design variables: fiber orientation angles of lamina.
The laminated composite plate under consideration is 64- ply laminate made of
graphite/epoxy. The fiber angles are integers varying between -90 and 90 ( 90 90 ) with
different degree increments in the laminate sequence.
In cases where the angle variation are continuous and discrete, the optimization results
are compared. In addition, the optimization methods used in similar optimization studies in
the literature and the results obtained under the thesis (based on DE and SA) are also
compared. The critical buckling loads are maximized for the factors of load case and plate
aspect ratio, and are compared with published results. Moreover, critical buckling load factor
has been investigated for the materials Boron/Epoxy, Graphite/Epoxy, Carbon/Epoxy,
Kevlar/Epoxy, S2 Glass/Epoxy, Fiberite/HyE9082A, S-Glass/Epoxy, E-Glass/Epoxy,
Flax/Epoxy, E-Glass/Polyster, Alfa/Polyester and Flax/Polypropylene. As a result, it has been
found that the parameters load, load ratio and plate aspect ratio are effective for optimization
of the critical buckling load factor that increases buckling resistance of laminated composites.
It is also observed that the DE method shows better computational performance than SA.