Keywords: topology optimization, tall buildings, skeletal frameworks, hybrid genetic algorithm, optimality criteria method.

Modern tall buildings are getting more and more irregular and complex in geometry. The structural design of such complex structures poses a challenging task to even experienced engineers. Without a formal optimization method, the search for the most cost-efficient structural topology as well as the optimal element sizes of practical building structures is generally a complicated and time-consuming process.

In this paper, the optimal topology design problem of tall skeletal frameworks is tackled by a decomposition-based design strategy coupled with an efficient hybrid optimality criteria and genetic algorithm (OC-GA) method. A decomposition-based design strategy is devised to reduce the computational effort required for topology optimization of tall steel building frameworks. Solution efficiency can be improved by decomposing a large topology design problem into a sequence of smaller and more tractable sub-problems, which can be solved by a multi-stage optimization process. Since topology optimization is a kind of combinatorial optimization, where the possible design combinations increases dramatically with the number of discrete topology variables, a reduction in the number of topology variables in each sub-problem can greatly reduce the design space and result in quicker solution convergence.

Once the topology design problem is decomposed into a set of sub-problems, an effective optimization technique with discrete applicability is needed for solving each sub-problem of the topology optimization. This paper presents an efficient hybrid genetic algorithm for design optimization of structural topology and its element sizes of tall skeletal building frameworks. The so-called hybrid OC-GA method combines the global search capability of the GA and the efficiency of a rigorously derived optimality criteria (OC) method. While the GA is particularly useful in the global exploration for optimal topologies, the OC method serves as an efficient local element sizing optimizer for topologies created using the GA. Given the topologies generated by the GA, the deterministic OC operator exploits the Karush-Kuhn-Tucker (KKT) necessary conditions in resizing efficiently all individual structural elements to their optimal sizes. The local search OC operator also facilitates local improvements in a given topology by first identifying inefficient elements having sufficiently small cross sectional sizes and then removing such inefficient redundant elements from the topology. Through the local search OC operator, GA-generated topologies can be locally improved and the overall efficiency of the hybrid OC-GA method is thus improved.

A twenty-storey planar skeletal framework has been used to illustrate the effectiveness of the decomposition-based design strategy using the hybrid OC-GA method. Results indicate that the multi-stage OC-GA approach is computationally much superior to the single-stage OC-GA approach. The decomposition based multi-stage OC-GA method provides a promising tool for optimizing both the topology and element sizes of practical tall building frameworks.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £140 +P&P)