M i c h e l a T u r r i n

PERFORMANCE-ORIENTED ARCHITECTURAL GEOMETRY FOR LARGE MODULAR STRUCTURES

The design process of large structures is explored by focusing on structural morphology and cladding system, based on performance evaluations of design alternatives during the early phase. Particularly, thermal comfort of spaces covered by large structures, such as atria or large roofs, is often problematic. The current research supports the design of such structures to passively improve the underneath thermal comfort in order to reduce the need for imported energies by using on site renewable resources. The main focus is given to modular structures and a range of various tessellations is explored. The integration of kinetic/adjustable modules is investigated to provide adaptiveness to seasonal and daily changing climatic factors.

Parametric design and genetic algorithms are used to support the design process. Parametric modelling has the capacity for creating design alternatives while managing complex interrelations since it represents components’ characteristics and their interrelations in a network of hierarchical dependencies. Different values of the independent parameters generate geometrical design variations. Due to the large solution space of most of parametric models, systematically evaluating the performances of each generated variation is not possible without the integration of further computational supports. Help is given by using search methods combined with the generation of large sets of alternatives. For this purpose, the research uses ParaGen, a genetic algorithm based optimization system developed by Peter Von Buelow at Michigan University. 

The research activities are promoted by Prof. Sevil Sariyildiz (Delft University of Technology, Faculty of Architecture, The Netherlands), supervised by Dr. Rudi Souffs (Delft University of Technology, Faculty of Architecture, The Netherlands). Among the consolidated collaborations, a few are mentioned: with Dr. Peter von Buelow (Michigan University, Taubman College of Architecture and Urban Planning, USA ), Dr. Axel Kilian – PhD MIT (Princeton University, USA). Special acknowledgements are given also to Prof. Massimo Majowiecki.