Anisotropic composite structures are widely used in aerospace, marine, civil, and biomedical areas of
engineering due to their advantages, including excellent specific strength, resistance to fatigue and
damage tolerance behaviour. Multiple crucial slender structural components of aircraft, automobiles,
buildings designed to withstand various loads are modelled as composite beams, thus it is very
important to understand the structural behaviour of composite beams and to investigate the
mechanism that causes their static deflection.
In this thesis mathematical models describing static deflection of composite beams and composite
beams resting on elastic foundations are investigated using both analytical and semi-analytical
methods based on Euler-Bernoulli and Timoshenko beam theories. These models for the static
deflection of composite beams, presented by a system of coupled ordinary differential equations with
corresponding boundary conditions, are rigorously derived. The nature of the governing equations
depends on the particular problem. For example, the homogeneity of equations is affected by the type
of applied loads, while the coefficients of the governing equations are determined by constant or
variable stiffness properties of the beam and elastic foundation. In order to obtain closed-form
analytical solutions for the problem, coupled governing equations are rewritten in a compact matrix
form enabling direct integration to uncouple unknown variables. Closed-form solutions are presented
by formulae computationally more efficient compared to commonly used numerical methods such as
finite difference or finite element methods, providing deep insight into the mechanism and physics of
the static displacement of beams, and quantifying the role and importance of model parameters.
Subsequently, semi-analytical techniques, namely the variational iteration method and the homotopy
analysis method, are used to predict the static behaviour of composite beams.
The presented analytical models are fast and computationally efficient which can be utilised during
the preliminary design stages. The derived results can be utilised as benchmark solutions to assess the
accuracy and convergence of various analytical and numerical methods.