Thursday, June 24, 2010

Humble Parametrics

Blog entry by: Dominik Holzer
Design computation and in recent years particularly parametrics and scripting have opened up a vast array of possibilities for creative minds to approach design problems in unprecedented ways. Taking advantage of the rule based approach computational design offers, many architects have done what they do best: exploring the possibilities of new tools at hand by testing the grounds in a playful and investigatory manner.

Most commercially available CAD support prior to the advent of parametrics or scripting has been mainly deterministic - either focusing on the replication of manual drafting tasks more efficiently, or assisting with visualisation.  Parametric design and scripting on the other hand has now proven its worth in allowing architects, engineers and others to explore their concepts in an open-ended manner. They can define dependencies in a model that then orchestrate the interplay of criteria and constraints that govern geometrical behaviour and help to assess performative responses.
Since the early days of parametric design, the constraint based approach to modelling allowed users to generate multiple highly differentiated and complex-looking variations for morphological experiments.  The problem with such experiments was that many of those conducting them seemed to be satisfied with producing seductive topologies that lacked deeper engagement with the ‘why’ aspect; thereby missing out on the potential of rule-based design for more substantial and useful applications. This resulted ever more similar looking and repetitive geometrical outcomes that included various themes of ‘smooth’ variation and repetition with little regard to materiality, constructability, or even the purpose of the parametric exploration. In that sense parametric design became a style rather than a design approach, which is best exemplified with Patrik Schumacher’s use of ‘Parametricism’.

At the same time, there are those designers who see the potential for using computational design for addressing specific design problems as they occur in everyday practice – being it ‘tame’ (well defined) or ‘wicked’ (more complex, open ended) ones. The use of parametrics in this context is often less motivated by geometrical experimentation for formal reasons, but by creating an array of informed design variations that relate to a specific goal or context. These variations could be used to optimise the configuration of building components based on clearly quantifiable constraints, or they could be applied to search for satisficing geometrical solutions that address physical building performance.   An example of the former would be the versioning of a stair and its railings, whereas an example of the latter would be the design of a sun-screen that helps to optimise daylight levels while avoiding sunlight-glare. This ‘humble’ use of parametrics can apply to a complex topological surface as much as it applies to a ‘standard’ modernist building. It is not a question of style, but rather a question of adequacy of using a the right design tool for a specific purpose.

In recent years software developers have searched to extend the parametric features of their tools to also allow for more fundamental geometrical configurations to become rule-based. Autodesk has merged some of the environmental analysis functions from its ECOTECT tool into a parametric modeling framework based on REVIT via VASARI, we now also see multiple plugins for McNeel’s Rhino/Grasshopper environment to link to structural (SALAMANDER/KARAMBA) and energy analysis (GECO).  

The examples shown in this blog entry stem from an undergraduate class I ran at SIAL in Melbourne who’s design brief included the implementation of the same project brief for a client at two climatically diverse locations in Australia. Students tested their design skills when combining parametric optimisation with environmental analysis to optimise the configuration of their proposals. 

No comments:

Post a Comment