Читать книгу Rethinking Prototyping - Группа авторов - Страница 95

The here described canopy is part of an ongoing project-based research that seeks to challenge structural typologies inherited from modernist practice and nineteenth century design procedures still prevalent in our built environment. In a series of projects we questioned structural system purity, which is often favoured by engineers for the sake of repetitive detailing and simple analysis, but neglects the architectural design approach and site-specific requirements. Our aim, in contrast, is to look forward and design both novel structural systems and the necessary computational tools and procedures (Bollinger et al., 2008).

System homogeneity is useful in vast structures with mono-functional infrastructural use and large spans (Billington, 1985). Migrating structural purity from the mono-functional into the complexity of architecture might, however, prohibit a dialogue between the load bearing and the space forming.

The canopy is one of several projects in which the purity of a structural typology went through a computationally driven evolution and alteration beyond structural optimisation. All projects have in common that their structure is not a superimposed preconceived typology but the result of a negotiation process balancing multiple criteria. The aim is to embrace further programmatic, formal and architectural considerations into the design of structures.

The history of projects is accompanied by the ongoing development of digital design tools. Available tools and procedures have a major impact on how we design. Bollinger + Grohmann Ingenieure constantly urges for close collaborations of architects and structural designers and therefore develops tools for seamless data flow between generation and analysis (Preisinger, 2012). It is the link between computationally generating structures as geometrical objects and their analysis and evaluation that requires permanent improvement. Once this connection is defined by the architects and the structural designers, computational power can be instrumentalised to not only represent but to generate variety and a multitude of possible solution (Bentley, 2002). In several projects (Fig. 1), we worked with structure generating algorithms that create vector-active systems with counter-intuitive placement of elements. The algorithm starts with many versions of systems made from stochastic distribution of structural members. They become the objects of analysis only after the generation and hence do not carry the burden of a preconceived typology.

Fig. 1 A series of projects (Top: Sphere Deutsche Bank, Pedestrain Bridge Reden, Sphere Deutsche Bank. Bottom: Music Pavilion Salzburg, Skyline Frankfurt) investigates novel arrangements of elements within vector-active structures. Based on computationally driven design processes structure emerges from an evolutionary process.

The process of analysis, evaluation, selection and feedback into a subsequent generative iteration becomes the driver of novel approaches beyond well-known systems. The evolutionary principle gradually migrated into the computational design tools of architects and engineers within recent years. The design process of the canopy shows the need for this development. The shortcomings identified during the development of the project pointed towards the need for an advanced tool development that finally lead to Karamba, a plug-in for Robert McNeels & Associates’ Grasshopper.