From Form Analysis To Morphology Generation

Rationalization of architectural design has been a long-term pursuit in the history of human architecture.The exploration of numerical generation is embedded in the exploration of architectural design rationalization.This thesis traces the historical evolution of the application of mathematical tools in architectural morphology generation.Mathematics,including geometry and proportional order,played an important role in classical Western architectural morphology generation.In the modern architectural movement,geometry and proportional order became rational tools for escaping the constraints of historical styles.From the 1960 s to the 1990 s,contemporary architect Peter Eisenman’s exploration of form generation under the influence of Western formalist trends to the early exploration of digital architectural morphology generation has become a key figure in driving contemporary digital architectural generation design.Taking the evolution of form analysis,form generation and morphology generation in the history of Western architecture as the main research thread,this thesis compares the transformation of the role of mathematical tools applied in different stages,taking into account the background,theoretical thinking and research tools of each period.Mathematical tools in architectural design have gone through the process from auxiliary generative tools to active generative tools: that is,from mathematical measurement,number ratio analysis to numerical generation,and architectural design has begun to realize the transformation from result rationalization to process rationalization.First,this thesis traces the pursuit of geometric and proportional order in Western classical architecture,and the use of modal grids and orthogonal axes in modernist architecture.The exploration of form generation in Western formalism,including Heinrich Wolfflin’s perceptual analysis,Rudolf Wittkower’s symbolic analysis,Colin Rowe’s structural analysis,and Peter Eisenman’s textual analysis,is traced,illuminating the evolution from formal analysis to formal autonomy in contemporary Western architectural thinking.This thesis focuses on Eisenman’s early exploration of digital architectural form generation,which is based on generative ideas in philosophy and linguistic theory,exploring autonomous operations based on the internal logic of form and even the use of digital tools to generate architectural forms.In the 21 st century,a new generation of pioneering architects represented by Greg Lynn,based on Eisenman’s early exploration,along with complexity thinking,interdisciplinary cooperation and the wide application of computer technology in the field of architectural design,generative design has gradually moved away from the "vacuum" characteristic of the mere pursuit of formal autonomy to a more holistic and the "morphological" characteristics of the integrated design.The transformation from form generation to morphology generation can be summarized into two development paths: external morphology generation and logical dynamic generation.External morphology generation is a semi-black-box process due to the fixed characteristics of structure and material and the obvious external characteristics of emerging geometry,and semi-visibility of results,i.e.,there is a certain control and vision of the final result.In contrast,logical dynamic generation is based on the rules and logic set by computer programming to automatically perform a large number of operations and generate unpredictable forms.The results generated by different computational models have no obvious class characteristics,and the unknown nature and dynamic diversity of the results belong to a black-box generation process.Although the rules and functions are written in a way that frees the hands from manual operations,the strong professional barriers do not completely free the brain,and they are also called "symbolic AI" methods.Finally,this thesis sorts out the current intelligent building design methods with machine learning and deep learning as the main technologies.Compared with the previous symbolic AI methods,the development of machine learning and deep learning makes the user more in line with the intelligent requirements without the corresponding programming skills or strong knowledge threshold,and from the manual operation stage,automation stage to the intelligent stage,the designer has gone through the process of freeing his hands,semi-liberated brain and liberated brain.Based on the controllable and innovative synergy of intelligent generation methods,the collaborative participation of human and machine and the systematic nature of the process,three generation directions of architectural stylization,performance design and urban space scaling have been developed.In the field of architecture,digital technology,artificial intelligence technology and complex science have jointly promoted the change of architectural design paradigm,and a new theory of architectural form generation has been formed in this context.In this thesis,the history and theories of digital architectural design are sorted out with the clues of form analysis,form generation and morphology generation,and this thesis is useful for clarifying the evolution and future development direction of architectural design.