Desktop version

Home arrow Education

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Form finding—design as search

Frei Otto understood that form wasn’t a neutral container. Through complex material experiments at the Institute of Lightweight Structures in Stuttgart, Otto was able to allow for the “self-organization” of form in design. This is a technique that had been pioneered by Antonio Gaudi in the physical calculation of the vaulting of the Sagrada Familia in Barcelona in the late 19th and early 20th centuries. From hundreds of experiments Otto was able to derive optimal form by allowing the forces of stress to reach a state of equilibrium. He demonstrated these principles through the use of soap bubbles, where surface tension forces the thin membrane of the bubbles to conform into spherical arrangements. When introducing external elements, soap bubbles would self-organize to determine an optimal solution to the problem of force. The emergent minimal surface is a physical phenomenon obtained by the optimization of stress under interpolation constrains.

Form finding suggests that the act of designing becomes a search for an optimal solution. Adopted from evolutionary biology, the notion of the fittest design is considered validation criteria for an architecture that has been conceived as a puzzle, where one solution presents the ideal response to a problem of performance. The practice of material computation, from a parametric perspective, inherently declares the variables associated with structural stability as the key field to be evaluated and changed to reach to an optimal structure. It hasn’t been until the advent of digital simulations, with techniques such as spring-relaxation or subdivision algorithms, that the practice of form finding could include architectural variables beyond structure, such as inhabitation areas, light exposure, egress paths, etc. These are criteria that experiments in material computation would have missed, as physical materials are agnostic to the series of overlapping objectives of an architectural agenda. In physical form finding, the hierarchical imperative is gravity and self-support, digital form finding allowed the negotiation of material parameters with a larger pool of architectural concerns. What emerges out of the accumulation of design parameters in a model is that inevitably, there will be criteria that counteract one another at the moment of evaluating one form of performance; for instance, as exemplified by John Frazer in one of his lectures in 2014,19 the width of the hull of a vessel will affect its speed and its potential inhabitation at opposite ends of the variable’s domain. In such a case, the designer cannot just rely on form finding but must first decide what criteria is most relevant for the brief in hand. Or, as has been presented by philosopher and Professor Steven Shaviro as an argument against methodologies of self-organization:

No self-organizing system can obviate the need for such a decision, or dictate what it will be. And decision always implies novelty or difference—in this way it is absolutely incompatible with notions of autopoiesis, homeostasis, or Spinoza’s conatus. What we need is an aesthetics of decision, instead of our current metaphysics of emergence.2"

While suppressing concerns like budget or ethical labor practices can yield exuberance, form is always an act of decision making. Design is the selection of a series of parameters and criteria over others. A naïve form of computation suggests that there is indeed an optimal form, one that might be able to consider all the available parameters and objectively determine the apex of a fitness landscape, but as some criteria are privileged, others will inevitably be disregarded.This form of optimization demarcates a boundary between the object and the environment, implying that the optimization can only be operated over the object of governance. The object becomes an island, able to optimize all its internal components but speciates from any further couplings.

If we focus our attention at the fitness landscape, and observe design as a probabilistic event defined by all the degrees of freedom enabled by a parametric equation, we can identify that such landscape does not offer all real design solutions. Such a fitness landscape needs to be understood as a system of expectations.

<<   CONTENTS   >>

Related topics