Algorithmic Generation of Complex Space Frames

In this research we want to investigate the generation of complex space frames by the means of algorithmic processes and estimate the impact and the relevance for architecture and structural engineering. The applicant has gained the required experience for this project via the development of different projects. The principle feasibility has been proved and different approaches have been asserted. The contemporary methods of structural design are based on a systemization of structures, which has been initiated by Karl Culmann in the middle of the 19th century. In this way structures with a high degree of effectiveness, which have a high degree of regularity, can be developed. Today the desire for free form design and therefore the need to design irregular structures can be viewed in built projects, as well as at architecture schools. To develop irregular structures that feature optimized load bearing capacities, the usage of existing design methods are limited due to an approach based on reductionism. Here is a lack between the requirements and claims on the part of architecture and civil engineering and the tools available. Non regular structures that have intrinsic load bearing capacities are complex. To evolve complex structures it is necessary to regard not only the whole system, but even more the relationship and the interaction of the system`s particles. Modern Hard- and Software allow calculating and evaluating load bearing capacities of almost every structure in short time. It is possible to analyse a great many of alternative solutions of a system and develop optimized systems iteratively. Therefore structures, that feature emergent load bearing capabilities, can be generated algorithmically and are not a priori based on a fixed system. The goal of the proposed research is to analyse new and innovative approaches to develop irregular and at the same time effective structures. Its practical aim is not only the generation of optimal structures, but the integration of construction principles that ensures an inbuilt practicability.

In the project "Algorithmic Generation of Complex Space Frames" basic approaches were explored to design new types of bearing structures. This way bearing structures are not a priori defined in one step by taking systemization as a basis, but a process is established, in which a structure with emergent load bearing capacities crystallize from a large solution space. Strut and Tie structures persist of single beams and allow the construction of large span roofs and bridges as well as towers. These are mostly drafted as frameworks and their load bearing capacity is hereby per definition reduced to pressure and tension. Geometry of these is following the triangle of forces arranged as triangles. This leads to a constrained canon of different types of framework systems that have an intrinsic degree of regularity. Structural systems are actually much more complex than displayed by such systematic model. In each framework structure bending moments appear. Modern hard- and software allow detailed and realistic simulation of a structure bending moments and its effective load bearing behavior. Practically all structures are independent of underlying load bearing principles and structural sense predictable. Here is a gap between the possibilities to evaluate complex strut and tie models and the tools needed to design these. The main emphasis of this research project was to develop stand-alone software to close this gap. The result is GENTs (Generic Exploration and Navigation Tool for structures). Using GENTs structures do not have to be predefined. Within defined boundary conditions and in short time a large amount of random structures are generated and calculated. These are in an iterative process improved by utilizing principles that are taken from evolution like selection, recombination and mutation, until an effective structure is gained. The results display highly irregular structures, load bearing impacts get complex and are not explainable at first glance. In studies the possibility to extent existing design vocabulary was shown. Furthermore lighter and more effective structures can be designed, in particular when dealing with irregular boundary conditions. The resulting structural systems are able to make by efficient utilization of material resources not only an important contribution to sustainable development, the possibility to define structural system procedural features a new quality for design processes and for cooperation between architects and structural engineers.