Influence of Initial Mesh Topology on the Optimal Structural Design of Steel Space Frame Structures

Abstract

Space frame structures today are a common choice of load bearing system for achieving long spans with minimal interruptions of the floor plan beneath. On top of that, the versatility of space frame structures to conform to any shape makes them particularly interesting today, especially in the context of free-form geometry which is becoming ever more common. In structural design, and especially for space frame design, the creation of a structural model is quite a labour-intensive process, and it is highly beneficial to automate the structural model generation. Furthermore, the design of such structures can benefit from an exhaustive preliminary design space investigation. Thus, this MSc thesis deals with the parametric design, engineering and optimization of space frame structure typologies based on different initial surface discretization of the input free-form surface geometry, and different topological relations of space frame top and bottom layers. These topological relations are based on Conway operators most relevant for the structural patterns occurring in space frame structures specifically, dual, kis and ambo. These operators are always applied on an initial or seed meshing of the desired free-form surface, to create different space frame layouts for them to be compared by their structural performance, primarily in terms of mass. The scope of initial meshing options is kept to tri, quad, and skeleton-based quad meshing. These three different types of mesh options, combined with the three possible Conway operator options, constitute the main nine combinations for each case study example.
In essence, every space frame design, due to the linear and geometric nature of the structural elements (nodes and bars seen as points and lines), can be considered as a literal structural translation of the final desired free-form shape. This free-form shape is always tessellated or discretized in certain configurations. The aim was to gain more insight into how the initial tessellation affects the behaviour of space frame structures as well as how the process of optimization of such structures is influenced regarding the initial tessellation. To gain insight into this influence qualitatively and quantitatively a parametric tool was developed to conduct case studies. The tool allows for the generation and cross-section optimization of various space frame structures based on an input surface. This parametric tool was developed using Rhino, Grasshopper, and karamba3D structural analysis plugins for grasshopper.