Doctor of Philosophy
School of Civil, Mining and Environmental Engineering
Sharafi, Pezhman, Cost optimization of the preliminary design layout of reinforced concrete framed buildings, Doctor of Philosophy thesis, School of Civil, Mining and Environmental Engineering, University of Wollongong, 2013. http://ro.uow.edu.au/theses/3916
This thesis proposes an explicit methodology for optimising the cost of the preliminary design layout of framed Reinforced Concrete (RC) buildings. The design of the column layout for RC buildings was studied first, followed by the optimum location of columns for the RC structures. The next step was to optimise the layout including the location of the columns and the rectilinear shapes for the plan layout and the rectilinear shapes proposed for the RC buildings.
In order to formulate such a methodology a model was developed to represent the mathematical relationship between the costs of concrete, longitudinal steel, shear steel, and formwork, and the number and size of the spans and the shape of the plan. Such a model illustrates how variations in the layout affects the cost elements, and how each cost element varies when the layout changes.
A unique mathematical relationship between the design layout and cost elements was formulated for structural systems such as continuous beams, plane frames, 3D frames, flat plates and beam-slab floor systems, in order to study different responses towards variations in the layout. This is a novel approach that can easily be used to optimise the design of various types of large RC structures and also account for constraints imposed by the design standards.
This approach makes use of cross sectional effects such as the positive and negative bending moments, and the shear and axial forces along the structural members for each structural system to derive a novel cost function as an alternative to traditional cost functions for optimising the layout of RC structures where design variables are used. These novel functions can be used to optimise the cost and layout of RC ii structures, because they are represented in a new space that takes the cost elements and layout design variables into account.
A combined approach was adapted to deal with a new structural optimisation problem, so an automated technique for optimising the preliminary layout of framed buildings with rectilinear patterns is presented. This method supports all types of rectilinear buildings (also known as orthogonal or iso-oriented) plans, including rectangular frames where the number and size of the spans and the shape of the plan can be variables. To that end, the knapsack problem was used as a basic applied combinatorial optimisation problem.
Ant Colony Optimization (ACO), as a robust metaheuristic, was used to solve the combinatorial optimisation arising from the structural optimisation problem. Numerical examples for each structural system are presented to demonstrate the robustness and practicality of the methodology and algorithms.