Computational design of reciprocal frame structures: Incorporating force variables into design process

Abstract

Form-finding methods are mainly driven by forces and the state of equilibrium, terms architects do not typically engage with. Although form-finding is not a sheer creative act, it can be conceived of as more than just composing forces in a designerly way of speaking. Therefore, the way architects attempt to steer forces can encapsulate different design perspectives on form-finding methods. Architects usually use parametric design tools during the design exploration, afterward they evaluate the structural behaviour of their designs as a separate step. Even though physics-based engines allow for considering structural behaviour during the design process, they lack extended geometric space in which designers explore unexpected results that fall outside the defined design space. This thesis develops a comprehensive workflow that investigates a form through computational design and making processes while bringing structural efficiency forward. The research focuses on finding a reciprocal frame (RF) structure form at the convergence of geometry, structure, and material. RF structure is studied due to the intrinsic dependency between its geometric configuration and structural behaviour. RF structure behaviour mostly relies on its geometric configuration, unlike other spatial structures for which the material and overall form determine the structural behaviour. Any change in the design variables resonates in the whole geometric configuration thus the structural behaviour changes accordingly. Therefore, top-down design approaches in which the final form is pre-determined have been studied extensively. Recent studies of RF structures focus on improving its structural behaviour but they were realized as a separate step from form-finding process as well. There is a gap in investigating bottom-up approaches in which the structural reciprocity of the form enhances the structural behaviour simultaneously. Therefore, the following research question is asked: How can we create RF structures where interdependent design variables of structural reciprocity inform the structural behaviour? The research method consists of three main processes: 1) geometric analysis, 2) physical experiment, 3) digital form-finding. The workflow is predicated on the data flow of three processes provided during the research. Initially, geometric analysis is realized by developing a definition in an algorithmic design environment. The algorithm evaluates the geometrical efficiency of RF in terms of forming the same surface with the least material. The analysis results show that diamond or rhombus geometries can form an RF NURBS surface by requiring fewer linear elements. Therefore, rhombus geometry is investigated in the physical experiment process. The computational making process gives rise to the design of a Volumetric Frame Unit (VFU) which essentially increases the structural stiffness of RF. VFUs are joined together based on a rule-based design system to explore different RF configurations. The VFU's geometry provides an input for the digital form-finding process. Digital form finding process is undertaken through 3-Dimensional Graphic Statics (3DGS) which is an intuitive structural form-finding method. The intuitive part stems from its ability to design forces of the form visually through polyhedral shapes. Therefore, the VFU geometry is abstracted as a polyhedron which is a rhombic dodecahedron. The resulting form diagram of 3DGS which works in pure compression is transformed into an RF structure. Transformation into an RF is realized by following analytical geometric operations. Finally, the outcome is evaluated using structural analysis software. The results emphasise that the way elements assemble, and compose a volumetric configuration enables the structure to work under compression forces along with bending forces. This thesis contributes to the computational design field by developing an extended workflow incorporating the force variable into the design process, which is not prevalent for architects and designers. The research has demonstrated that the comprehensive design workflow in which geometry, structure, and material are intertwined paves the way for exploring outcomes with structural primacy. Consequently, structural primacy mostly associated with mechanical terms can be integrated into the design process which is recognized as a creative act to give rise to feasible and aesthetic forms.