Papapavlou, A.; (2008) Structural Evolution: a genetic algorithm method to generate structurally optimal Delaunay triangulated space frames for dynamic loads. Masters thesis , UCL (University College London).

Preview

PDF 13920.pdf Download (10MB)

Abstract

An important principle in the architectural design process is the quest for the optimum solution, a quest which is in this study structurally motivated and necessarily computationally oriented given its high complexity in nature. The present research project suggests an evolutionary algorithm that draws its power from the literal interpretation of the natural system's reproductive process at a microscopic scale with the scope of generating optimal Delaunay triangulated space frames for dynamic loads. The algorithm repositions a firm number of nodes within a space envelope, by establishing Delaunay tetrahedra and consequently creating adaptable optimised space frame topologies. The arbitrarily generated tetrahedralised structure is compared against a canonical designed one, whilst several experiments are conducted in order to investigate whether -and to what degree- the genetic algorithm method is appropriate for searching discontinuous and difficult solution spaces or not. The results of this comparison indicate that the method proposed has advantageous properties while being capable of generating an optimum structure that exceeds statically the performance of an engineered tetrahedralised space frame.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item