Parametric Design

Topic Point 5: 

What is parametric design? Is it the "style" of the future? Is it the key component to an architectural movement that “finally offers a credible, sustainable answer to the drawn-out crisis of modernism that resulted in 25 years of stylistic searching” as Patrik Schumacher proposes in his Parametricist Manifesto? What can computational design offer to a profession already in the midst of a transition, or "crisis of representation" as was repeated during recent thesis presentations? The shift from designing in a two-dimensions to designing in a three-dimensional environment, at least within academia, has questioned many fundamental understandings of how architecture should be represented. Many students produce plans and sections, cut from their 3D designs as opposed to designing in plan and section, while at the same moment there are successful practicing architects that only work in 2D and rarely touch a computer. Is there a lack of understanding and discourse across a "generational gap" in the profession that is only exacerbated by the exponential advancement and appropriation of technology in architecture? How does the implementation of parametric design "play out" in regards to a "formal aesthetic" or "style", mass customization (with regards to both design and production), and/or the scripting of an architecture that will actually respond to parameters, either within the currently static design environment, or even mechanically, after its construction?

For a great discussion already underway and a strong inspiration for this prompt go to:

cmuarch2013.wordpress.com/2010/05/06/parametricism-style-of-the-21st-century/

Adapted from an essay for:

California College of the Arts

Course: Architectural Analysis

Instructor: Michael Bogan

Spring 2010

By: Sean Stillwell

Parametric design software is just that; software. It is a tool to achieve whatever the designer sees fit within the limits of that tool. It is a medium through which design concepts can be explored, formally or otherwise. Of course, tools lend themselves to "styles" or "aesthetics" but it has been, and always will be up to the designer to either understand and utilize any given tool for both its strengths and weaknesses, or to not fully grasp the utility of that tool and thus be subservient to its tendencies. Designing within a set of parameters, or rule sets, is not a new phenomenon in architecture. What has changed with architecture's appropriation of computer-aided design is the exponential computational power utilized through various parametric software platforms. This allows for greater complexity within the algorithm, the embedment of scripts within parent scripts, and even for the iteration of scripts themselves according to a host of additional parameters, all outputting formal, organizational, and/or even conceptual iterations at an accelerated rate. While parametric design software has made available the exploration and production of exciting and complex geometries foreign to architecture, its potential to revolutionize the profession is not a conceptual one. Instead, this software offers architects the ability to embed design environments with information and/or conceptual frameworks that allow for a mutually determinant relationship between architecture and its context, the ability to advance the mass customization of both design and production, and the ability to script building system controls that actually react to given parameters mechanically, in real time, throughout the life of the building. 

In the article Animate Form, by Greg Lynn, the author begins with the idea that architecture “produce[s] drawings of buildings and not buildings themselves [and therefore] is involved with the production of virtual descriptions” (Lynn. 10). Many of these virtual descriptions are problematic to Lynn in that they are conceived of in stasis, devoid of force and motion. A dynamic alternative is a design strategy that is described by “static coordinates”, where “the virtual force of the environment in which it is designed contributes to its shape” (Ibid. 10). When multiple vectors representing time and force are included design space they can influence the topology of a surface and its ability to occupy multiple positions simultaneously in a continuous form.  Lynn calls this “animate design”, where “motion and force are co-present at the moment of formal conception,” and goes on to describe an “architecture [that] can be modeled as a participant immersed within [the] dynamic flows [of the virtual force of the environment]” (Ibid. 11). By giving “weight” to objects or points that influence the object’s form in this dynamic space, the resulting architecture would be in dialogue with the initial conditions of that design space. This would create a mutually determinate relationship where the resulting architectural intervention has a direct affect on the conditions that influenced the formal design. Lynn's description of "animate design" is a conceptual framework in which parametric design software can operate by scripting parameters into a design environment where those environmental parameters are in constant negotiation with the architecture produced.

In the past time laps techniques have been used to imbed art and architecture with motion. Often form influenced by time and force has been represented as an overlay of numerous static frames as depicted in Marcel Duchamp’s Nude Descending Staircase No. 2. Another design strategy involves “sequential formal operations recorded in the building through colors, alignments, imprints, additions and subtractions” (Ibid. 13). Lynn contrasts these methods of imprinting time and motion on architecture with that of designing in an environment populated by forces. Fixed relationships to program, and the consideration of gravity as a simple vertical force are concepts of stasis that can be limiting to design when simply reinforced or contradicted. Designing in an environment with time and force can transform these assumptions into numerous possible relationships that are lost in the idealized stasis of our current 3D modeling software. “Independent interacting variables can be linked to influence one another through logical expressions… from which either a single or a series of instances can be taken” (Ibid. 14). These types of relationships allow multiple forces to affect multiple points differently, which in turn start to influence each other and the form that they define.

Marcel Duchamp, Nude Descending Staircase No. 2

Found at: www.beatmuseum.org/duchamp/nude2.html

The relationships Lynn refers to are key concept when designing with parametric software. By setting up complex relationships between points based on real and/or conceptual information that are either adjustable, or adjust themselves based on specific criteria, a dynamic system develops that can incorporate design influences like force and time. Parametric Software allows for architecture to explore and understand relationships based on integral and differential calculus that can imbed not only a changing value, but also the rate at which that value is transposed into formal, organizational, efficiency, or conceptual strategies. 

Transposition into a form from these malleable relationships can be arbitrary and problematic if both conceptual and pragmatic filters are not rigorously applied. This contemporary ability to loading form with force and time with the incorporation of parametric software challenge architects “to understand the appearance of these tools in a more sophisticated way than as simply a new set of shapes. … gradients, flexible envelopes, temporal flows and forces” (Ibid. 17). The "globular" forms often associated with parametric design are not characteristic of a style or movement in architecture other than the ability to produce them through the advances in both building construction and computational design. The ability to produce these forms offer a sort of freedom in the guise of possibility, but the success of the architectural concept should be judged by how the resulting design performs given the architects intent. Simply embedding form with motion without a conceptual and boarder-line functionalist rigor can produce an architecture that has traded stasis for irrationality.

This is one critique that can be given to the Parametricist's Manifesto, by Patrik Schumacher, partner at Zaha Hadid Architects and co-director of London's A.A. Design Research Lab, and its descriptions of the Kartal-Pendik Masterplan, Istanbul, Turkey, 2006 by Zaha Hadid Architects. While "irrationality" may be a little harsh, the preliminary issue I will take with the manifesto is the attitude taken in the utilization and explanation of parametric design software. The choice of describing parametric design as a "style" is incorrect. Even calling this technology a movement, which is preferable to "style," is problematic. Schumacher defends his use of "style" as a descriptive term: 

“...architectural styles are best understood as design research programs conceived in the way that paradigms frame scientific research programs. Thus, a new style in architecture is akin to a new paradigm in science: It redefines the fundamental categories, purposes, methods, and innovative course of a coherent collective endeavor.”

While the exploration of parametric design should be an “innovative [and multi-directional] course of a coherent collective endeavor,” paramedic software is a tool. Just as economists have utilized scripting in Excel, the ability perform complex algorithms on multiple data sets and organize results into legible visualizations for analysis, architects can begin to overlay variables such as environmental aspects of site, embedded energies of materials, and/or organizational factors into computational scripts. Priorities, determined by the architect, are placed on each variable and then formal outputs can be analyzed. Understanding the implications of the results and how they are formed is the key element in differentiating parametric design from a movement. Even when a dynamic data set produces emergent (and exciting) results through parametric software, a rigorous understanding is necessary, not only of which specific parameters produced a given results, but also of what their implications are. The selection of initial inputs, the determination of how they are scripted, and a critical analysis of the result(s) are all choices. The trajectories enabled by those choices are delineated by the interpretation of a concept, and the means by which the architect can and should remain in control of the design process. 

That is not to say that the time and resources that Zaha Hadid's office, and more specifically Patrik Schumacher, have invested into understanding these tools are not valuable to the architectural community, but the implementation of parametric design in the Kartal-Pendik Masterplan, for example, expresses the problems associated with either a lack of control, a lack of understanding, or a lack of representation and explanation of using “active tools” like Maya’s hair “diagrams.” In my limited experience with Maya hair “diagrams,” as Schumacker compares to the wool form finding experiments of Marek Kolodziejczyk, the results are at best uncontrollable, nonhierarchical, and un-nuanced. Maya animation platforms are use to produce an effect used in animation, not a rigorous diagram that can be transposed into an architectural and/or urban logic. This particular tool offers a handful of variables that only amplify or deaden the effect of static cling between follicles and control the speed of the process. With a developed script, methodology, and a real understanding of what variables affect which aspect of what result, parametric software should embed each thread with a prescribed logic based on specific parameters. This is a very different experiment then simply running a series of splines through a filter and choosing a particular result that may or may not have matched a preconceived form. In this project there was an opportunity to create a hierarchy amongst the connective tissue between the “adjacent contexts” that could have been rigorously controlled and nuanced within a script, much more so than Maya’s hair “diagrams” are capable of. 

Zaha Hadid Archiects, Kartal-Pendik Masterplan, Istanbul, Turkey, 2006
Maya hair-dynamic simulates minimised detour net. The path network was thus generated with a digital woolthread model. The set-up registers the multitude of incoming streets and bundles them into larger roads affording larger parcels. Found at: www.patrikschumacher.com/Texts/Parametricism%20-%20A%20New%20Global%20Style%20for%20Architecture%20and%20Urban%20Design.html

Zaha Hadid Archiects, Kartal-Pendik Masterplan, Istanbul, Turkey, 2006
Masterplan: Hybrid detour net & deformed grid, Final Urban lay-out of streets and urban fabric. Found at: www.patrikschumacher.com/Texts/Parametricism%20-%20A%20New%20Global%20Style%20for%20Architecture%20and%20Urban%20Design.html

While this specific example shows the weaknesses in the limited understanding of a specific tool, the utilization of software outside of the “normal” architectural toolbox is a conceptual step in the right direction. The automotive, aerospace, and even sporting goods are industries that already embed design software with variables like wind resistance and gravity to analyze the performance of their designs. Ecotect and other similar environmental analysis software also begin to address some of the opportunities of embedding contextual information within a design environment, but its analysis is far from a dynamic, mutually determinant design environment. With the incorporation of analytic plug-ins such as Ecotect into open source parametric engines such as Grasshopper or Rhinoscript for Rhinoceros 3D, for example, the vacuum that architects have been designing in could become obsolete. This is exactly why the research at well known firms like ZHA is so important. With larger projects, larger budgets, and greater resources, their research has the potential to further our ability to design with parametric software and develop the knowledge to construct the subsequent products.

Wilson is taking its soccer balls to the next level in aerodynamics research using CD-adapco’s STAR-CCM+ software on Windows HPC Server 2008. 

Found at: www.prweb.com/releases/2010/07/prweb4253844.htm

The conceptual digestion of an overwhelming cache of real time data, the immediate demand and gratification of the “online,” and the schizophrenic nature of personal, mobile technology can be facilitated through a parameterizing of relevant data banks and utilizing these real-time inputs as field conditions within design environments. In designing within a contextual, albeit virtual, environment, gradients have become the means of describing data-scapes of environmental site conditions and occupation that in turn have augmented architecture's understanding of boarder conditions and transitional spaces. This is not only in regards to a "Semperian" notion of "inside and outside," but also with regards to human comfort, programmatic layout, and systems integration. These subsequent gradient fields call for the customization of our built environment and the means to construct that environment at all scales, from large "formal moves" down to individual components comprising possible emergent conditions. Schumacher describes parametric design allowing architecture:

"Systematic, adaptive variation, continuous differentiation (rather than mere variety), and dynamic, parametric figuration concern[ing] all design tasks from urbanism to the level of tectonic detail… Architecture finds itself at the mid-point of an ongoing cycle of innovative adaptation – retooling the discipline and adapting the architectural and urban environment to the socio-economic era of post-fordism. The mass society that was characterized by a single, nearly universal consumption standard has evolved into the heterogenous society of the multitude… The task is to develop an architectural and urban repertoire that is geared up to create complex, polycentric urban and architectural fields which are densely layered and continuously differentiated.”

I agree with the analysis that society at large, and more specifically the architectural community has moved from Modernism into a pluralistic, "heterogenous society of the multitude" where there is no one defining style or movement. In contrast with modernism, there is a conceptual shift from the ability to control or subjugate nature to a methodology looking to exploit environmental factors at a given site responsibly and improve conditions when possible. The current "retooling" and "adapting" of the discipline that Schumacher is directly referring to is, of course, parametric software, but the tool is one of many means and not in itself an end. Parametric design, devoid of its own agenda, will eventually gain the acceptance of the larger profession and of its clientele through economic means of efficiency where design and production of mass customization in a "post-fordism" era advance in a manner that is mutually determinant. The ability to design through scripts provides the efficiency to design customization that can be at odds with the means to produce it. If that design is deemed economic by another factor such as its material efficiency, for example, then that motivation may be such that the means of production are improved. An innovation in a particular means of customized production may also generate the economic motivation for design to innovate and harness it. 

RECURSIVE GROWTH | … / February 2008

“Self-similar subdivision system” 

Exploration : MARC FORNES / THEVERYMANY™
Design, Research & Computation: Marc Fornes

* V3.0

[THEVERYMANY “series” have yet no pretention to be smart as biomimetism or aiming toward new ecologies… there are based on a very simple and straight forward oldschool empiric lab approach trying first to understand existing mathematical/geometrical models as explicit and ordered sets of instruction, learn to replicate them through coding and finally within that process strategize in terms of design…

RECURSION in mathematics and computer science, is a method of defining functions in which the function being defined is applied within its own definition. The term is also used more generally to describe a process of repeating objects in a self-similar way. For instance, when the surfaces of two mirrors are almost parallel with each other the nested images that occur are a form of recursion. (ie wikipedia),

RECURSIVEGROWTH – Generation 1 to 7 (2,4,6 omitted) – is based on a coming back to “Recursion” (ie previous tests on subdivision, etc…) as the ultimo model of periodicity – why periodicity after many non-linear approaches? simple: the last series based on aperiodic tilling or replication were used because of high repetition within its model – therefore whenever one is approaching the concept of repetition its ultimo and endless quest is maximizing it…]

Found at: theverymany.com/exploration/08-recursivegrowth/

Another key conceptual issues I take with designing within the dynamic, parametric, 3D environment described above is that the incredible flexibility is lost once the production phase is reached. While certain conditions can be iterated and developed as forms, organizations, etc., during design, variables, adjustable or reactive to real time data streams that affect architecture, become static once some if/then statement is satisfied or when a desirable form is reached and the design is then "baked," to use the Grasshopper term. All of the inputs are frozen, whether that may be the reality of the inputs or not. Construction documents are printed and given over to the contractor and construction is underway. What was conceived of as a dynamic architecture becomes a static form when built, whether understood conceptually as a field or object. Of course, there are building systems that remain reactive to human comfort like air conditioning or lighting (an on/off switch or operable blinds), but the very nature of parametric software begs the exploitation of an architecture's ability to react to those same design parameters after construction. While this may seem like science fiction, there are many practices that are exploring just this type of research at all scales. Future Cities Lab is one such practice that is exploring this trajectory and have helped develop a plugin for Grasshopper called Firefly. This plugin interfaces with Arduino an open-source processing language and micro-controller that creates a link between a 3D design environment and a mechanical prototype. These studies, while currently at a relatively small scale can be extrapolated to offer possible applications at a facade or entire building scale. Again, it will be the economic factors that will either allow or hinder the development of this application of parametric design. When the dynamic parameterizing and movement of components becomes cheaper or more responsible than a greater density of material, or some sort of post-construction energy generation technique can outweigh the initial cost/benefit ratio of its implementation, then the innovations of this particular niche become practical. 

Energy Farm in Seoul, Korea [2nd Prize Winner - Seoul Performing Arts Island Competition 2005]
The “Urban Energy Farm” proposes a complex intermeshing of variable interacting agents. Three interacting fields of the ground, the space frame and the sky-pins compose a differentiated yet synthetic landscape that is capable of finely controlling and generating a series of varied environment. Though a synthetic set of parameters, this proposal works to create difference at the micro scale of inhabitation. These fully responsive sets of site systems respond, fluctuate, transform and produce according to environmental or energy needs of program and inhabitation. The parametric tools of design become energetic agents in the conception and activation of urban space.
Project Credits: Jason Johnson, Nataly Gattegno with Anthony Viola, Beth Haber. Future Cites Lab.

Found at: http://www.future-cities-lab.net/index/?cat=6

In my brief 2 years in a Graduate Architecture program, I have been introduced to numerous tools and techniques for design without always understanding the conceptual implications of the forms that can be produced.  Developing a familiarity with the tendencies and limitations of design methodologies as well as the analytic rigor to critique both process and artifact are necessary parallels and even foundations to any design studio. Cultural issues of pragmatism, responsibility, and conceptuality consistently challenge emergent theory and technology and the spatial consequences of using any tool for generative design should be analyzed through specific lenses of contemporary and past culture to determine both their relevance and weaknesses. Parametric modeling, for example, can offer the ability to absorb existing mathematics and scripting into digital design environments where field conditions of data-scapes can be generated and analyzed. It allows for the ability to shift scales from systems to subsystems with seamless transitions through mass customization and can facilitate the design, construction, and/or the operation of dynamic architecture that will continue to mechanically interact with its context, its occupation, or other factors to be determined by the architect. But this technical ability is not a worthy "conceptual framework" or "movement" in architecture. It is a tool, a means to explore how these new possibilities can implement an architectural concept.

Bibliography of sorts:

Lynn, Greg.  Animate Form. New York: Princeton Architectural Press, 1999.

cmuarch2013.wordpress.com/2010/05/06/parametricism-style-of-the-21st-century/

blog.marcantonioarchitects.com/architectural-quackery-at-its-finest-parametr

patrikschumacher.com/Texts/Parametricism%20-%20A%20New%20Global%20Style%20for%20Architecture%20and%20Urban%20Design.html

nzarchitecture.com/blog/index.php/2010/09/25/patrik-schumacher-parametricism/

pushpullbar.com/forums/showthread.php?10250-Parametricism-Is-it-really-the-next-great-architectural-style

grasshopper3d.com/forum/topics/refactoring-parametricism

lebbeuswoods.wordpress.com/2010/07/15/rendering-speculations/

theverymany.com/exploration/08-recursivegrowth/

future-cities-lab.net/index/?cat=6