Research On Design Methodology Using Generative Systems As Operative Logics

Throughout the endeavor of architectural research, architects have been attempting to set up methodology systems for architectural design. In the era of information, digital design possesses the advantage of constructing design methodologies following the technological development, by its quantitative and systematic usage of algorithm knowledge. Generative system design is a systematic process that generates diverse outcomes by using algorithm logic and regularized programmes. It could lead to a design methodology which combined bottom-up research strategy and top-down control of system operation. The generated projects have unpredictable possibility, continuous differentiation and parametric controllability, thus integrating analysis, concept, design manipulation and optimization. Generative system design gains its theoretical validities in principle fundamentals of Gilles Deleuze’s analogy of "rhizome" and related researchesThe core research target of this thesis is design methodology using generative systems as operative logics. Generative systems will serve as base logics in this research, while the ultimate objective is to summarize a methodological system using generative system in architectural design, by manipulating projects. Therefore, after briefly introducing the related theoretical researches, the author focuses on three design examples:pavilion design, urban design of Merwehaven Transformation and architectural design of Merwehaven New Ferry Station. The operation mode of different generative systems separately in these three examples would be illustrated in detail, with the specific design processes from background analysis to final architectural form of each project introduced. By researching on these design examples, four crucial phases of design using generative systems as operative logic would be pointed out:identification of input and conditions; algorithm modeling of operative logics; generating computational topology; form making. In the afterwards development of these projects, optimization tools like genetic algorithms, which also belongs to generative systems, and other computational approaches will play an important role, leading to the improvement of building performance.This thesis reflects on generative system design after studying other similar examples. Two aspects would become the main concerns of this design methodology. There exists logical gap between computational topology and final architectural form, which could result in the lack of control on the architectural presentation. Meanwhile, the lack of consideration on the construction ability in design process would bring the problem as absence of optimization in terms of construction. To make progress in the future research, the author should work on adding up top-down control by purpose on generative systems themselves and architectural fonn making, as well as designing generative systems with consideration for construction.The whole thesis contains about 46 400 words,157 pictures and charts.