Computational analyses of dynamics of architectural space

Abstract

Although space is the central element of architecture, questions of space are hardly addressed in computational research in architecture. On the other hand, current mainstream practice in computational design research in architecture tends to focus on efficiency of architectural procedures, entailing optimisation of form, structure, performance, data management or workflow, etc. Such focus utilises computation to handle quantitative data of tangible properties in architecture. However, architectural space is filled with abstract qualitative properties, one of which is its dynamics. Dynamic properties of architectural space have been discussed in diverse disciplines from diverse perspectives since the nineteenth century, but it is only in the past decades that some of the theories are revisited due to discoveries in neuroscience. Such reappraisal of past theories by new technologies anticipates further rediscovery of qualitative properties of architectural space, such as spatial dynamics, that have been investigated largely through speculative descriptive methods using phenomenological approaches. Hence, the research explores the idea of architectural space as dynamic field structures by reexamining theories since the nineteenth century in multiple disciplines, and develops a system of computational inquiries to investigate dynamics of architectural space. The computational procedures produce visual spatial data that are analysed and calibrated in comparison to the past studies of architectural space based on descriptive methods. The correlations observed between the two approaches substantiate potentials of the computational approach as ways to study abstract properties of architectural space further.