Form and Morphogenesis
in Modern Architectural
Contexts
Domenico D’Uva
Politecnico di Milano, Italy
A volume in the Advances in Media,
Hershey PA, USA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail: cust@igi-global.com Web site: http://www.igi-global.com
Copyright © 2018 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark.
Library of Congress Cataloging-in-Publication Data
British Cataloguing in Publication Data
A Cataloguing in Publication record for this book is available from the British Library.
All work contributed to this book is new, previously-unpublished material. The views expressed in this book are those of the authors, but not necessarily of the publisher.
For electronic access to this publication, please contact: eresources@igi-global.com. Names: D’Uva, Domenico, 1975- editor.
Title: Handbook of research on form and morphogenesis in modern architectural contexts / Domenico D’Uva, editor.
Description: Hershey, PA : Information Science Reference (an imprint of IGI Global), 2018. | Includes bibliographical references and index.
Identifiers: LCCN 2017026110| ISBN 9781522539933 (hardcover) | ISBN 9781522539940 (ebook)
Subjects: LCSH: Architectural design. | Space (Architecture)
Classification: LCC NA2750 .A615 2018 | DDC 729--dc23 LC record available at https://lccn.loc.gov/2017026110
This book is published in the IGI Global book series Advances in Media, Entertainment, and the Arts (AMEA) (ISSN: 2475-6814; eISSN: 2475-6830)
Chapter 19
DOI: 10.4018/978-1-5225-3993-3.ch019
ABSTRACT
After nearly three decades since their first appearance in architectural practice, digital design tools are increasingly pervasive in nearly every aspect of the profession and throughout the building life cycle, from project development to construction administration to demolition and recycling. While an integrated approach to building information management is becoming the key to winning projects, the creative attitude of an earlier generation of computational designers is being quickly replaced by new tools and protocols geared toward achieving efficiency targets and boosting profitability. The author reflects on the evolving nature of the digital practice and the potential for a new generation of architects to resolve diverging aspirations towards creative freedom and efficient use of resources. The chapter draws on a few experimental projects by the author that combine traditional design tools with computational tech-niques to explore a direct correlation between building form and energy performance while forging a new vocabulary for sustainable design.
INTRODUCTION
The traditional boundaries of architecture have been greatly expanded by technological advancement in detecting, measuring and visualising the energy forces acting on the built environment—forces that were largely invisible to the architect’s eye until two decades ago. Over the same period of time, the so-called ‘digital revolution’ brought to architects the tools to design and construct formal systems that are potentially capable of responding to these underlying forces with unprecedented precision and resolu-tion. These new powerful tools came to fruition in the midst of an environmental crisis of extraordinary proportions, giving design professionals the opportunity to play a central role in solving the energy chal-lenge. The effect of the new digital regime, however, has been a gradual drifting towards extravagant and often wasteful design propositions. By and large, the flourishing of a new formal vocabulary, enabled by digital tools, rarely translates into buildings that perform better.
Energy/Form/Rule:
Experiments in Energy Form-Finding
Simone Giostra
With energy efficiency and rational use of resources becoming the overriding concerns in both new construction and retrofits, architects have been gradually marginalised in the design and construction process, as they have failed to provide responsible answers to the environmental crisis engulfing the planet. The resulting leadership vacuum has been rapidly filled by myriad of new professional figures— energy specialists, feasibility consultants, construction managers—who have all made their way into the construction cycle, often relegating the architect to the role of concept designer producing fancy renderings for the client.
The Energy Challenge
Europe has undertaken an ambitious roadmap towards the reduction of greenhouse gas (GHG) emissions by 2050 and, more recently, the commitment to COP21 targets is producing the first obligations at the EU level to achieve significant cutting of GHG emissions by 2030. Most of these targets entail efforts in the built environment, especially in cities, which are recognised as the drivers for change. The reduction of energy use and GHG emissions poses an intrinsically multi-scale and cross-disciplinary challenge—one that appears well suited to the skill set of the architect—yet it is mostly tackled by deploying a plethora of isolated and highly technical solutions.
After more than two decades of environmental policies in place, achievements in energy savings related to the built environment have been disappointingly modest. In fact, most results have been achieved due to factors outside the architectural and the urban design fields. For instance, reduction of GHG emis-sions occurred as a result of technological innovations in the mobility and energy sectors, such as more efficient energy plants, an increase in clean sources within the energy mix of networks, green innovation at the scale of building components and more efficient appliances (“Progress on Energy,”2016).
The projects presented here are part of a decade-long effort by the author to recast the on-going debate on sustainability in the built environment from a pre-eminently architectural position. Very dif-ferent in scale, program and location, they all explore the correlation between form and energy in its various manifestations. Collectively, they start to identify a specific, measurable relationship between geometry—the traditional domain of the architect—and performance, particularly in the area of energy efficiency and sustainable use of materials in buildings. These projects are the result of concrete com-missions, either by a private client, a research institution or a competition authority; as such, they were developed within a specific timeframe and budget, using the limited resources typically available to an emerging architecture practice today.
Generational Divide
Advanced CAD and parametric tools, which were the exclusive domain of the aeronautical industry and a few large engineering firms only a decade ago, are now available to young practitioners to develop and control complex shapes. Similarly, digital simulation tools that predict the environmental performance associated to a particular building form—such as airflow, daylighting and sun radiation levels—are becoming increasingly more accessible and easier to use, providing an intuitive and inexpensive sketch tool to designers. These tools can have a profound impact on the way we design buildings and cities, since they provide invaluable support at a very early stage of the design process, when most decisions are made, as opposed to entering the process at a much later stage, when the designer is less willing to accept changes to a consolidated design. Also, they come at a much lower cost, since they replace the kind of quantitative advice traditionally offered by specialised engineering firms.
Needless to say, these tools are extremely controversial within more traditional academic and profes-sional circles. They empower a new generation of young designers and disrupt the current teaching model, which is still largely based on the artistic tradition of the Ecole de Beaux Art. Outside the school, they are set to transform professional practices as well, as fresh graduate students experiment with new forms of collaboration that bypass the traditional apprenticeship model—a source of cheap labour on which the success of most corporate firms is predicated. Perhaps for the first time since the gradual specialisation of the profession started in the US during the 1950s, young professionals have the tools to win sizeable commissions without relying on a vast network of consultants or the resources of a large office.
The design establishment, of course, has been trying to disable the impending revolution. In aca-demic circles, these new digital tools tend to be relegated to ancillary functions of primary intellectual processes that still promote ideas over tools. Larger architectural offices also struggle to find a place for computational strategies within a reliable workflow, particularly in the early stages of design. As a result, they prefer to devote digital resources to describe and represent formal solutions generated by more traditional methods.
A quick look at the evolution of digital design culture in recent years might shed some light on the conflicting, often confusing nature of the transformation at hand.
Gateway to Another World
A claim frequently heard from older colleagues, both at school and in the profession, is that new forms of digital practice—that is, using a machine in the artistic process—stifles creativity and generates anonymous architecture.
In fact, the problem of design is not, and it never was, one of creativity—of enabling the mind to formulate new formal constructs, that is, of ’coming up’ with ideas—but quite the opposite: the creative process desperately needs parameters, limitations, some kind of intellectual friction in order to operate. Unchecked, the mind is capable of conceiving the wildest shapes, none of which would actually turn into architecture.
The first set of limitations comes in the form of a tool enabling thoughts to take shape in some in-telligible way. Ideally, a useful design tool would limit the range of expression to only that which can be eventually built. If the tool is too restrictive, the result will be conservative or conventional design; conversely, if the tool is too loose or irresponsive, you will end up with wild propositions that cannot be built. If architectural drawings are the anticipation of the act of building, then tools are the guardians of the act of anticipating, deciding which line/form/structure has a right to exist in a drawing.
Many older modelling and visualization tools, such as 3ds Max and Cinema 4D, knew no boundaries, since they were created by the film animation industry precisely to ‘unbound’ the imagination of the designers and to create an imaginary world that needed to exist only on screen: “your gateway to another world” is the promise by a leading developer of 3D software. The evolution of graphic interfaces only meant an increasingly smother transfer mind-mouse-screen that effectively eliminated many, if not all, limits to creating free forms—making the results largely irrelevant to the purpose of building.
The fact that an early generation of architects were being introduced en masse to these tools in the mid-1990s by elite architecture programs at Columbia University in New York, SCI-Arc in Los Angeles or the AA in London, speaks to the opportunistic, disingenuous relationship between architecture and the graphic software industry. By and large, this first generation of ‘digital’ designers were responsible for stretching, once again, the boundaries of what was considered architecture—and for raising much of
the opposition to the so-called ‘blob architecture’ that is still felt in academic circles today, along with a lingering suspicion toward any new digital tool that has come since.
From the Command Line
Digital design, however, has many strands. At the opposite end of the spectrum, early CAD tools offered a great deal of resistance in the form of a reduced number of operations, none of which included unforeseen or unimaginable results. A ‘command line’ implies a master able and willing to spell out orders: here every shape is the result of explicit instructions given in a specific sequence, using a mediating protocol that requires training and some ability. Similar to the visualisation tools discussed earlier, CAD software was developed for the engineering industry, not architecture, and presents its own kind of limitations. For one, it forces the designer to a level of precision that finds no application in architecture, particularly in the early stages of the project. Also, it does not allow for any tolerance, since it demands that each line should be placed in Euclidean space without ambiguity or hesitation, with the snap function marshalling any wandering line to its designated place. Incidentally, Building Information Modelling (BIM) software belongs to this second lineage of digital design tools.
An intelligent 3D model-based software that involves functional and relational characteristics, BIM represents a more pragmatic and conservative response by the industry to the same disruptive forces transforming all levels of design. In fact, BIM tools are designed to enhance productivity and ultimately profitability—one major player in the market incites architects to “use BIM architectural design software to win more work and retain clients” (“BIM Software,” 2017)—at the expense of innovative and risk-taking approaches to design. Contrary to a generalised perception that BIM software should empower the architect and foster design innovation, the author is convinced that it will regiment the creative process in favour of delivering normalised, predictable (and profitable) results.
In fact, BIM software represents the antithesis to the experimental processes pioneered by early digital artists and the hacker culture that infiltrated many artistic fields over the past 40 years—first electronic music, then video art, interactive design and gaming—and slowly percolated into more traditional design fields such as architecture, with the introduction of graphical algorithm editors like Grasshopper and Processing. The projects presented here are emblematic of this early experimental phase, when algorithms were used to propel as much as disrupt the traditional design practice.
Generative Tools
Because of its disruptive potential, a conservative majority still perceives the digital practice in opposi-tion to analogue modes of design, such as hand drawing or model making—a futile distincopposi-tion at best, serving the entrenched interests of an older generation. And it’s not merely a fight for self-preservation. Some of the most exciting digital tools making their way in the profession are generative in nature, that is, they open up design opportunities that become apparent only to those who practice. As with any other craft, there is no verbal substitute for a digital practice. Most decision makers, in our schools as well as in most traditional industries, simply lack the digital skills to appreciate first-hand the generative potential of these tools.
For centuries, architects have been using scale models to predict the performance of buildings by applying materials and techniques that replicated actual constraints. Physical models, however, cannot test a design solution for structural integrity, as commonly accepted before the discovery of the so-called
square/cube law by Galileo in 1638. For over 300 years from the first publication of the ‘Two New Sci-ences’, we did not have a reliable analogue method to test structural integrity of buildings in the early phases of design. This is particularly striking if we consider, as Reyner Banham noted, that the history of architecture up until the end of the 20th century is largely a history of space-enclosing structures (Banham, 1969).
Remarkably, intuitive computer simulation tools provide designers with the unprecedented ability to test early design concepts for structural integrity, together with many other performative aspects of the project, effectively overcoming a centuries-old limitation. And there is more: the introduction of accessible parametric tools, such as Grasshopper a decade ago, allows the designer to use rules and algorithms to generate forms, resuming the tradition of form-finding that consumed the best minds of an earlier generation of architects—including Frei Otto’s experiments with lightweight structures and Gaudi’s analogue force models.
Rather than a fictitious opposition between digital and analogue models, then, what’s really at stake is a dramatic shift in recent years from form-making to form-finding.
Today, energy considerations are supplanting structural integrity as the main parameter in designing a building. Thanks to advances in computer technology, we are the first generation of architects with the tools to simulate relevant energy indicators from the very early design concept, using inexpensive applications run on our laptops. Of particular interest are tools that produce graphic output in the form of 2D- and 3D-color-coded diagrams, in some case projected directly onto the space being evaluated. The defining challenge for our generation is to do for energy what Frei Otto and Gaudi did for structure: shaping buildings using energy-related constraints—that is, energy form-finding.
Invisible Forces
The problem with parametric design tools is that they require explicit instructions to operate; in other words: they only execute orders. As any architect working on a design problem knows well, much of the creative process in architecture is based on what Malcom McCullough calls ‘intrinsic information’, that is, information that is embedded in the ambient and comes to fruition less through focused attention than by situational awareness. Creative work does not always involve deliberate thought; a skilful practice, tools as props, habituation—all play a cognitive role in ‘coming up’ with ideas. As he puts it, “A great deal of knowledge is inarticulable, especially when in use. In music, sports, or many other expertise, you can do things you cannot explain” (McCullough, 2013, p. 65).
While digital tools cannot replace the architect’s mind in formulating a design concept, they can be very helpful when dealing with information that does not fall within the visible spectrum. In one of his seminal writings, Buckminster Fuller famously declared: “[…] Forms are inherently visible and no longer can ‘form follow functions’, because the significant functions are invisible” (Fuller & Snyder, 2009, p. 359). He was referring to natural forces, as well as to material properties that are not detectable by senses or experience, since they result from manipulation at the molecular level that are invisible to the naked eye—yet have a great impact on the built form. Environmental analysis tools can provide critical insights into these invisible functions by widening the architect’s gaze in areas of knowledge outside the spectrum of visible light.
There are obvious advantages in giving form to these invisible forces, as they play an increasingly larger role in the built environment. Architects typically resort to highly technical solutions for compli-ance with ever-stricter energy codes— ‘green gadgets’ that come in the form of sophisticated mechanical
systems, super-insulation materials or expensive glass treatments—so that they don’t have to question a consolidated formal language. Conversely, formal solutions that directly address these invisible forces at a structural level can dramatically improve the performance of buildings by reducing heating and cooling loads, fostering daylighting and natural ventilation, and generally lowering energy demand.
Additionally, a building form that is the result of a form-finding process can manifest information regarding the ambient—prevailing wind direction, solar radiation levels, air flow or pedestrian traffic—in ways that are intuitive and do not require mediation. A classic example of the architect’s disconnected design approach to the new energy imperative are the many digital displays showing the amount of energy being produced by solar panels that are hidden away on the roof of buildings. This is particularly relevant in an age of mediated information: as we increasingly rely on screens, large and small, to retrieve useful information on our environment, embedding information in the persistent structure of buildings can have positive effect in learning to navigate our world without depending on a smartphone.
Beauty and Survival
Which bring us to a final question regarding the use of form-finding strategies and related computational systems predicting the behaviour of buildings. We understand that a form resulting from relevant forces might cope well with these same forces, so that if a building envelope is shaped based on solar radiation levels, for instance, it has a larger potential for energy generation than a building shaped after a crumpled paper bag. But how prominent should the energy radiation potential be among the many factors—such as program, context, budget, historic references or quality of interior space—contributing to the design of a building? The answer depends on the stage of human development in which you find yourself operating.
Cultural values govern the relationship between nature and human actions—a sort of protocol of en-gagement with our natural environment designed to improve the human species competitive advantage and, ultimately, chances of survival. Even abstract notions such as ‘beauty’, according to Denis Dutton, might be evolutionary determined, so that we consider beautiful that which enhance survival of the human’s genes (Dutton, 2010). By necessity, then, design criteria must be an evolving concept, as our collective success is continuously challenged by changing environmental conditions.
In ‘Collapse’, Jared Diamond argues that with changing environmental conditions, societies face the challenge of identifying which cultural values can be sustained and which ones are no longer appropriate to the new set of conditions (Diamond, 2005). He writes about the choice of Greenland Norse to stick to Christian identity values— for instance by refusing to adopt habits and technics from the indigenous Inuit that were much better adapted to the environment—as the main cause of their extinction. He describes how societies on the verge of environmental collapse, such as the Rapa Nui civilisation, stubbornly cling to—and sometimes even intensify—the very same practices that are the root cause of their demise. Erecting large ceremonial statues on Easter Island turned into an unsustainable practice toward the end of the 17th century, as it required a disproportionate amount of timber and human labour to sustain, in a context where sources of both trees and proteins were depleted. Statues became increasingly larger and more complex, therefore demanding more resources, precisely when resources became more scarce.
Confronted with the progressive depletion of resources and declining quality of our natural environment, we continue erecting monuments to our minor gods. In the eye of future generations, the irresponsible use of resources to serve the extravagant formalism of some of today’s most prominent architecture will bring to mind the excesses of a collapsing civilisation. And, if history is any indication, cultural conser-vatism is not what will get us back on track. On the contrary, this is a time for vigorous experimentation and some serious debate on what we collectively can and cannot afford.
News From the Trenches
The author’s interest for computational design strategies started in the mid-2000s with the early con-vergence of environmental analysis and parametric design tools—two areas of digital practice that have been widely expanded in recent years, but that were rarely combined as part of a consistent design process back then.
A few experimental firms were already producing architectural structures with great formal com-plexity, such as Gehry’s Guggenheim in Bilbao or Libeskind’s Jewish Museum in Berlin -showing an exceptional level of geometric control and sophistication. The creative process behind these colossal projects, involving hundreds of highly educated professionals and large sums of money, however, was often very rudimentary. The design concept, by the architect’s own admission, is based on a rather tra-ditional, at times whimsical ‘visionary concept’ expressed by a paper model or a sketch. The implicit nature of the initial creative process is clearly at odds with the explicit computational language utilised to send instructions down the chain of command. The projects presented here are an attempt to unpack that first inscrutable artistic moment—to unfold that crumpled paper bag, so to speak—and open up the process to organically combine a century-old tradition with the new tools at hand. Or, to put it in some historical prospective: to reconcile what Sigfried Giedion famously called the “schism between architecture and technology”, generated with the separation between Ecole Politecnique and Ecole de
Beaux Art by Napoleon in the French academic system in 1804. (Giedion, 1967)
For the past 15 years, the author has been experimenting with hybrid forms of computational design and traditional form-making, combining various performance analyses with parametric definitions, to inform and support a creative design process. In the tradition of form-finding, the first project for a com-mercial complex in LinYi, China, explores the theoretical limits of an algorithm to control a building envelope exclusively based on constrains. The second project for the MOCAPE in Shenzhen supplements a form-finding process with human intuition by creating a feedback loop between analogue and digital domains. Mapping natural forces shaping the site is often the first act of the process, clearly illustrated in the project for Land Integrated Photovoltaic in Almeria, Spain. The grain and scale of the architec-tural structure often conform to these underlying maps, as unique and appropriate building components respond to each data point on the field.
Each one of the following projects explores the potential for parametric design to construct this one-to-one relation between environmental forces, building form and energy performance.
Case Studies
1. BIG GREEN in LinYi, China
Application: building’s envelope Parameters: daylighting
Methodology: parametric definition + randomised distortion
Jingya Corporation, a leader in high-end hospitality operating in the Shandong Province for the past two decades, retained SGPA for the design of a radical new building, bringing together hospitality, entertain-ment and business under a vast vertical forest (Figure 1). The overall design of the building is the result of code restrictions, including setbacks, and solar radiation analysis, in order to optimise its shape based
on maximum levels of daylighting on adjacent buildings. The most prominent feature of the project is a 35.000m2 green envelope acting as responsive shading system over the entire complex—a veritable vertical forest with sufficient scale to develop its own micro-climate by regulating moisture content, direct solar radiation, and temperature.
The green envelope has particular significance in the humid subtropical climate of LinYi, as well as in the exceptionally harsh conditions surrounding the site, including a contaminated river facing the building, high levels of noise and air pollution from nearby traffic, and the generally degraded condi-tions of ageing buildings in the area (Figure 2). In this context, the green facade acts as catalyst for the upcoming restoration of the river and the reintroduction of native plant species along its banks.
Figure 1. View of green façade. (©2017 SGPA)
The project presents two seemingly contradictory conditions from the very start: on the one hand, zoning regulations, a relatively modest budget and the rational distribution of program meant that the building would have only strait, vertical facades, resulting in a terraced roofscape and a stepping profile on the main elevation (Figure 3, right). On the other hand, the design intent was to give a fluid, monolithic aspect to the building (Figure 3, left), so as to mark its presence within a highly fragmented, massively scaled urban context.
The sun-shading green envelope provides a solution in the form of a gently curving surface enclosing the inner volumes and resolving any discontinuity within the stepping floors (Figure 4, left). In a sense, the interstitial space between inner volumes and outer envelope reconciles a utilitarian building with its ambitious environmental agenda (Figure 4, right). On a more pragmatic level, the varying distance between body and skin allows for geometric articulation of the surface, mirroring the changing require-ments for shading, daylight and ventilation, particularly accentuated across the main south-facing facade. A primary challenge of some interest in this context is the resolution of a large-scale, double-curvature envelope using modular, off-the-shelf components and the traditional methods of construction imposed by the client. At the scale of the component, individual sun-shading elements needed to be adjustable to specific local conditions, without loosing formal continuity at the global scale, which resulted in apertures modulated by vertical bands (Figure 5).
The double-curvature surface of the shading component is constructed by using a flat stainless steel mesh, without resorting to costly metal forming techniques, as the mesh allows for easy twisting by pushing or pulling its edges in opposite directions (Figure 6). The mesh provides an armature for climber plants to grow and support themselves, as well as partial shading while the plants are still growing. The inner envelope consists of a modular, unitised curtain wall system that contains planters for the climbers, plus a point-structure support system for cables and metal mesh.
Figure 3. Design constrains: Monolithic massing (left); Stepping roofscape (right). (©2017 SGPA)
Figure 5. Diagram of gill-like apertures. (©2017 SGPA)
The interior program’s changing requirements for daylighting control the degree and direction of each individual stripe (Figure 7). Additionally, each of the approximately 180 stripes is adjustable at each floor of the building, for close to 30 points along its entire length, bringing the total number close to 5.400 control points. The architect needed a robust protocol to control these points both locally, to ensure compliance with daylight requirements, and globally, to coordinate individual fluctuations within a larger scale of figuration that provided an intermediate scale between the vast building mass and the micro texture of the stripe. Most crucially, because of the evolving program coming from the client that kept changing while already producing fabrication drawings for the envelope, the parametric model needed to be responsive enough for any last-minute change to daylight requirements to simultaneously reflect on all fabrication documents.
Figure 7. Drawing showing modular flexibility of stripes (left); Detail (right). (©2017 SGPA)
The first step of the process consists of mapping prevailing activities across the building’s floors onto the corresponding facade elevation (Figure 8, top). The elevation is then colour-coded based on intensity of daylight required by each program, with a gradient that ranges from red (less daylight) to light yellow (more daylight). The daylighting requirement map is used to generated a surface, or topography, with a comparable operation to height-field command in Rhinoceros, based on grayscale values of the colours in the map (Figure 8, bottom). In the resulting surface topography, a low daylight requirement corresponds a depression or cavity, while a bulge or protrusion indicates a high value, with each relative elevation on the map indicating the degree of aperture for the shading element at that point.
The following step is one of particular interest in this context, as it suggests a dialectical relationship between design intent and computational logic. The initial topography contains large areas with little variation, failing to express the ability of the system to continuously adjust to local conditions. In other words, the grain, or resolution, of the variation dictated by the program alone is too coarse and needs resampling. Several methods are tested to introduce small scale noise and variation to the topography, without altering the amount of daylight associated to each interior space (Figure 9). In order to produce a randomised distortion of the topography, the initial base surface (a) is offset in both directions, with the offset distance indicating the maximum range of geometric variation (b). Lines are drawn between corresponding control points on the two offset surfaces (c) and a random point is picked along the length of the curve (d). Finally, a new surface is generated based on the newly created points (e).
Perpendicular lines are drawn from the base plane of the building’s facade and extended to the limits of the new surface topography (Figure 10, left). These lines are placed on a grid coordinated with floor elevations and partitions layout, so as to work both as control points and structural supports for the en-velope. Horizontal lines are drawn to connect alternating endpoints of structural supports lines to form a bracing system of diagonals. Finally, these horizontal lines are lofted to create individual surfaces (Figure 10, right) describing sun-shading cladding panels.
Similarly, a parametric model is used for studies on panel’s size, with complete control over each panel’s height and width (Figure 11). This way of sampling design options, using a set of geometric variables for visual examination, is a quintessential demonstration of new opportunities brought by parametric models that were simply not available with traditional design tools. Although it was not part of this particular project, the same 3d model produced here could be tested for solar radiation and daylighting analysis, bringing yet another performative aspect to bear on the decision-making process.
The final parametric model includes geometric properties for each system component, all related to the originating surface topography, so that for any modification applied to the topography—as a result of changes by the client or other circumstances—all other components of the envelope are automatically adjusted (Figure 12). A unique challenge of this type of models is to make geometric information that is relevant to the manufacturing and/or construction process available to each party involved, in both a digital format for expedient retrieval and hard copies for all legal proposes.
Figure 10. Panelization process. (©2017 SGPA)
Figure 11. Parametric variation of panel size. (©2017 SGPA)
Information is provided in graphic format, indicating exact location of each data point, as well as in tabular form as spreadsheet. Salient information includes, for example, the varying lengths of structural metal pipes supporting diagonal and vertical mesh cables (Figure 13). Similar documentation is provided for each and every component of the system, for a total of over 25.000 unique components for the entire envelope, virtually updated in real-time for each subsequent modification to the envelope.
2. MOCAPE in Shenzhen, China
Application: building’s structure Parameters: load stress diagrams
Methodology: analogue model + structural analysis tool
A second-stage entry for the MOCAPE (Museum of Contemporary Art and Planning Exhibition) design competition in Shenzhen, the project’s main concept is a necklace of interconnected public spaces along a meandering path, mimicking the organisational spatial sequence connecting the many pavilions in the classical Chinese garden.
Figure 13. Sample information output: Elevation (left); Modular lengths of pipe (right). (©2017 SGPA)
Rooted in local tradition, the project also experiments with a new kind of open space for Shenzhen—a space that redefines the public use of the street level and extends to three dimensions, vertically connect-ing the city to the new museum. The buildconnect-ing symbolises the link between the museum, the city and its citizens in the 21st Century (Figure 14).
A sequence of dynamic circulation paths and large gathering spaces carving into the solid volume of the museum is the generator for the building’s form and structure (Figure 15). In consideration of the city’s sub-tropical climate, the sequence starts at the street level with a spectacular urban-scale outdoor space extending throughout the building, unobstructed yet contained within its structure. The large cavity spreads horizontally to divide the lower administrative program from the museum space floating above the plaza. It also extends vertically, articulating the different programmatic components of the museum, while reaching a scale evocative of the natural sublime. The floating volume creates a flexible area of shaded landscape, allowing sea and land breezes to pass through the site and promoting a stack effect for natural ventilation, in a new model of sustainability for the tropics.
The design’s main feature is a network of connecting horizontal and vertical voids, defined by a ‘slice envelope’ trimming the regular structural grid of the building—an inverted topography (green surface in Figure 16) containing the transfer beams and columns that carry loads to the ground through large scale pier-like structures. The sequence of interconnected open spaces transforms the building into a
Figure 15. Design concepts: Circulation paths carving into a volume; Necklace of public spaces (©2017 SGPA)
porous and breathing organism, where the flow of visitors constantly intercepts the natural elements of sun, wind and the greenery at the ground level. Alternate circulation paths provide for radically differ-ent ways to experience the building: gdiffer-entle stairs rising from the cdiffer-entral atrium offer an extraordinary journey through the various sections of the museum and a convenient exit way from each level of the building. Alternatively, a wide ramp brings visitors along a meandering path through a constellation of special program that offers alternative activities located in event pavilions, providing moments of relaxation and enjoyment.
Initially, a regular structural grid satisfies the basic requirement for extensive, flexible exhibition space for a Museum that had no permanent collection at the time of planning. A grid of columns 12m off centre extends over the entire 120mx120m footprint, 7-story high structure. The subsequent placing of special program requiring column-free spaces—the large atrium at the ground floor with intercon-nected voids, as well as the many halls along the meandering ramp—triggers the design response that is of interest in this section.
The insertion of larger, column-less spaces within a regular grid would be typically achieved either by subtraction via Boolean operation—for instance in the competition proposal by OMA in 1989 for the Très Grande Bibliothèque in Paris—or by stacking, as in the Hanover Pavilion by MVRDV for Expo 2000. Either strategy implies an extremely complex, costly and highly inefficient use of transfer structure to resolve the abrupt interruption of load paths generated by such operations, as a complete and continuous path is necessary for loads in any building to travel from roof to ground. In both examples, a highly praised conceptual simplicity requires a muscular (and often hidden) structural response. In fact, the large structural spans resulting from such operations—and associated changing column grids between stories—typically require a disproportional amount of structural reinforcement in the form of either diagonals, horizontal trusses, or both, in order to collect and transfer the gravity loads to the fewer support points available below. The Citicorp building in New York is a classic example of such load transfer structure, using diagonals in a chevron pattern on the facades and a 3-story deep truss at the base of the tower to collect floors load and wind shear to the centre of each face where the mega-columns below are located. Additionally, load paths are not always predictable, as loads find ways to make it to the ground that are not necessarily the ones designed by engineers. In fact, loads would typically trans-fer through the stiffest route, so that the actual load distribution pattern across the structure can deviate substantially from design assumptions.
The conceptual model of ‘pruning’ (Figure 17) provides for an alternate trimming method to either Boolean subtraction or stacking in order to achieve discontinuous column grids between stories. Pruning refers to cutting away dead or overgrown branches or stems, especially to encourage growth. Sometimes forest trees are pruned to increase the size of the trunk for timber purposes. The practice entails targeted removal of unwanted branches following specialised techniques, ranging from selective removal of stems to hedge trimming.
In all cases, pruning requires a deep understanding of the particular tree or shrub to be pruned, as well as the proper season must be known. Also, the position of the tissue being removed relative to the overall canopy should be considered, so as not to compromise the general stability of the tree. Conversely, pruning might entail the removal of appropriate branches to make the tree structurally sound. Most interestingly, the selective removal of particular branches has the effect of re-directing growth lines, as plants respond to pruning with a burst of growth especially just below the cut. The intrinsic ability of plants to develop these energy lines along available paths shows similarities with load path behaviour in building structures.
The form-finding process entails the use of an analogue model (Figure 18) made with wooden sticks reproducing structural beams and columns at 1/500 scale. With the use of precision pliers, individual sticks are gradually trimmed and the stumps connected by diagonals to create column-less areas and vertical cavities. Initially, the process is based on the intuition of the designer operating on the analogue model, guided by the evolving spatial qualities that start to emerge from the removal process. It should
be noted that contrary to a Boolean operation or stacking, selective and gradual trimming allows for constant monitoring and evaluation of the resulting spatial effects. Also, similar to the pruning of trees, the space-frame reacts to trimming by developing alternative structural paths to re-direct the load along the new diagonal columns and toward unaffected areas of the grid.
In order to evaluate these invisible forces being shaped by shifting load paths, the designer uses a digital model reproducing the initial frame, plus any subsequent pruning. The model is periodically up-dated and evaluated by a structural analysis software for relevant stresses, strains and deflections of the entire structure, providing invaluable real-time feedback to the designer on the effects of each trimming (Figure 19). Because of its highly interactive nature, load path reconfiguration requires an iterative pro-cess where modifications to the physical model are continuously mirrored by the digital model, which in turn is constantly tested for load stress, providing structural validation to either revoke the trimming or suggest ways to continue pruning.
The final result of the process can be described as a complete and continuous slicing envelope (Figure 20) trimming the structural grid, generated by diagonal folding lines acting as load-transferring members. The slicing envelope also identifies actual building enclosures, generated as ruled surfaces by connecting metal studs at regular intervals to top and bottom beams. The resulting load distribution at the end of the iterative process is surprisingly homogenous. Despite the wide range of structural spans occurring throughout the building, the load diagrams along individual column lines present a high de-gree of uniformity.
The resulting architectural space is quite original, shaped by careful crafting and rooted on a solid understanding of the dynamic structural behaviour of the system (Figure 21). This is a shape that could not have been conceived by other design methods—and certainly not by a traditional form-making process. Most importantly, the final shape of the building entails a level of structural integrity that guarantees efficient use of material and, ultimately, economy of construction.
The project demonstrates a hybrid design work flow comprising of a digital analysis tool and an analogue form-finding method, connected by a digital 3D model into a feedback loop (Figure 22). The iterative process allows for incremental optimisation of the structure, as the designer’s intuition is
Figure 20. Slicing envelope. (©2017 SGPA)
gressively refined by analysis conducted on a trial-and-error basis, ultimately developing into a skill that neither a traditional form-making process nor a strictly form-finding protocol could have produced. 3. LIPV in Almeria, Spain
Application: land infrastructure
Parameters: solar radiation, slope gradient and land use Methodology: various computational techniques
Figure 22. Work flow diagram. (©2017 SGPA)
Figure 23. Project area in Almeria, Spain. (©2017 Landsat / Copernicus, Data SIO, NOAA, U.S. Navy, NGA, GEBCO)
The present concept for a demonstration of Land Integrated Photovoltaics aims at providing a cultur-ally acceptable model for large-scale photovoltaic installations across the European landscape, in the context of new energy policies by the European Union.
Located in Europe’s driest desert, Almeria has the greatest concentration of greenhouses in the world: the mar de plastico, or sea of plastic, extends over 22.000ha in the Campo de Dalias area, producing millions of tons of produce for European markets (Figure 23). Initially fuelled by abundant aquifer water, today the sector is threatened by severe water shortage, resorting to hydroponics, massive desalination and cheap labour in order to remain competitive. Almeria is the epicentre of southern Spain’s two billion euro a year agricultural industry; also, it is at the forefront in pollution of water and aquatic ecosystems, salination of groundwater sources and severe pollution by nutrients, as well as racial segregation of the cheap immigrant labour force and ensuing social unrest.
The premise of the proposal is that any remediation effort must re-establish a sustainable water man-agement regime by a combination of collection and conservation measures, efficient irrigation systems, and the use of wind and solar energy to power water purification and desalination plants. Considering that the vast majority of existing greenhouses were built without any planning, that water infrastruc-ture is non-existing and that roadways are merely an afterthought, the project engages in a radical re-configuration of land parcels as the necessary condition towards a rational approach to production and efficient use of resources.
Computational strategies utilised for devising the new land division, the subsequent allocation of new functions and the development of physical infrastructure are the focus of this section.
Preliminary evaluation: Areas of interest are selected based on sun exposure and other relevant fac-tors, using satellite imagery, GIS and historic data publicly available. Within the selected areas, existing infrastructures and other physical systems are surveyed, in order to identify opportunities of integration. Issues of visibility, economy of scale, synergy with additional systems and services, accessibility and maintenance, productivity and efficiency are evaluated before selecting one or more areas for develop-ment. Of particular interest is the slope gradient analysis (Figure 24, left), as it reveals a correlation between gradient, geometry of greenhouses’ footprint, water streams and canals, with the solar exposure analysis (right).
Simulation: The main tessellation logic is derived from a simulation of water flow lines, using agents over an accurate 3D topographic model of the area; a set of parameters—amount of water, attraction capacity, reaction to gradient, etc.—are tested in different configurations to arrive to an optimal distri-bution of water lines (Figure 25).
Optimization: In the next step of the process, water flow lines become attractors with the ability to bundle up and form a consolidated mesh of connectors parallel to the coastline, to guarantee connectiv-ity across the region, and perpendicular to the gradient, providing water distribution to the lots. The size and shape of the regions is determined by parameters associated to the attractors and calibrated to obtain a tessellation that is appropriate to the existing modes of agricultural activity, property structure, water distribution and transportation infrastructure (Figure 26, left). The resulting tessellation of the land allows for an effective and detailed evaluation of the relevant energy forces associated to each tas-sel (Figure 26, right).
Tessellation analysis: Environmental analysis software is used to produce a series of maps for quantitative analyses of relevant parameters. As a result, each new land parcel is defined by values of solar radiation, slope angle, surface water drainage, soil content, and other indicators of environmental
Figure 25. Water flow simulation (left); Interpolation of flow lines with contour lines (right). (©2017 SGPA)
performance. Colour-coded maps resulting from the analyses offer an explicit, intuitive reading of large portions of land, while retaining a level of precision in the underlying data-sets associated to the maps.
In particular, solar analysis (Figure 27, left) shows generally high levels of sun exposure, with areas of intense solar radiation on the slopes surrounding the low plain. Currently, these areas are not used for greenhouses, offering a great opportunity to extend irrigated farming to areas that are potentially sup-ported by solar energy. The map of slope gradient (Figure 27, right) showcases an obvious correlation with the solar exposure map, providing complementary information and qualifying the gradient of each tassel, so that appropriate typologies can be associated to specific slope inclination. It also visualizes the threshold for lots that are not suitable for construction, due to excessive costs for excavation (red in the map).
The size analysis of parcels (Figure 28, left) mostly serves as control tool in fine-tuning the tessel-lation algorithm; lots that are too small or too large are subsequently re-divided to fit a size range that is appropriate to the type of ownership and production economy in the area. Pre-existing land uses are obviously a crucial factor in determining the new roles and they were maintained whenever possible, in an effort to minimize unnecessary conflicts with established property rights. In the land use map (Fig-ure 28, right), parcels that fall within the boundaries of existing productive units are mostly allocated to new greenhouses.
Figure 27. Tessellation analysis: Solar radiation (left); Slope gradient (right). (©2017 SGPA)
Program allocation: A particular combination of parameters may indicate the ability of a given parcel to perform a specific task—that is, it reveals the “vocation” of a lot for certain land uses—including, but not limited to, its potential for energy generation. Slope or gradient might be an indicator of vulnerability to land erosion, or additional costs associated to earth retention. Similarly, the ability of the parcel to drain groundwater might be connected to salinity levels, or other factors that might affect its intended use. Specific typologies are developed based on these features (Figure 29), including:
1. Low slope/medium solar: areas with larger greenhouses, urban patches with dwellings and public program.
2. Medium slope/variable solar: areas with smaller greenhouses and individual dwellings. 3. High slope/high solar: areas with medium-size greenhouses and high-density solar panels.
Typology development: A specific building typology is associated to each family of parcels; design strategies are then translated into physical infrastructures, using modular system components with the ability to adjust to local conditions and to respond “point-by-point” to the originating data set. In all cases, the particular design response is triggered by the associated environmental maps. The resulting modular system—whether dwellings, energy infrastructure, or productive units—is engineered to integrate with existing networks and with additional systems components that might be appropriate to the site.
Typology A (Figure 30) addresses the most common type of greenhouse on flat ground; the new tes-sellation follows the logic of water flow lines and topography. The resulting subdivision re-configures the existing land parcels by providing for rational water and vehicular infrastructure. The new tessellation also allocates areas adjacent to main access roads and natural features to urban dwelling and social services.
Typology B (Figure 31) refers to areas with moderate slope and is based on a rational and economic approach to construction, using traditional retaining walls, minimum excavation and landfill, and a so-lar roof with flexible orientation so as to respond to the most appropriate orientation within a moduso-lar system. The tessellation provides for an efficient drainage and retention system, so groundwater is ef-fectively intercepted and stored in local tanks, as well as for vehicular access to individual greenhouses
Typology C (Figure 32) addresses the challenge of bringing irrigated agriculture to the hills, where the potential for solar power is the highest in the area. The proposal uses the well-established tradition of terraced farming, with the addition of diagonal roads that provide vehicular accessibility to individual greenhouses, while pedestrian circulation is facilitated by stair connectors. Retaining walls offer an eco-nomic way of building covered spaces for produce storage and for water tanks. The system can adjust to a wide range of gradients, transforming into a pure solar panel installation on the steepest slopes.
Site integration: The new landscape of Campo de Dalia presents a fully integrated system: groundwater drainage infrastructure—including natural recurring streams and artificial canals—weaves through the new fabric of greenhouses, creating spaces for social gathering and public activities. These flood areas become the new public space that is entirely missing from Campo de Dalia in its current configuration (Figure 33).
Figure 30. Typology A: Bird-eye view (left); Site layout (right). (©2017 SGPA)
Land parcels are efficiently divided by secondary vehicular roads and tertiary pedestrian streets into units of appropriate size to maximize agricultural production while promoting walking and cycling as modes of circulation. Modular greenhouses oriented toward the prevailing sun direction cover between 80% and 90% of total land area, exceeding the existing green housing stock. At the intersection of productive land parcels with flood areas and road infrastructure, greenhouses include an urban edge as interface for loading docks, administrative offices, storage space, and dwellings. Areas that are particularly suited to habitation—or that include pre-existing villages—are entirely allocated to housing and public program.
CONCLUSION
All 3 projects deploy heuristic methods that combine aspects of traditional form-making and digital form-finding, with varying degree of automation in the use of a design protocol.
Figure 32. Typology C: Bird-eye view (left); Site layout (right). (©2017 SGPA)
It should be noted that for MOCAPE and Big Green, the point of departure is a rather conventional tessellation geometry: an orthogonal structural grid in the former and a vertical banding of the surface envelope in the latter. In contrast, the LIPV project uses a higher degree of automation at the initial stage, as it deploys an algorithm to simulate water flow over a given topography, then a path optimisa-tion funcoptimisa-tion to produce the base tessellaoptimisa-tion for dividing the land.
From there, MOCAPE continues with a manual operation that uses intuition to gradually trim struc-tural components. Conversely, Big Green and LIPV each adopt a performance indicator—daylighting requirements and a combination of size, solar exposure, gradient and proximity to water, respectively— to computationally drive the transformation of base geometry and/or allocation of program components. Big Green also uses a random function to accentuate a certain formal aspect of the design.
In the next step of the work flow, the geometry resulting from either manual or computational ma-nipulation is tested using relevant indicators, structural integrity for MOCAPE and daylighting levels for Big Green. Although LIPV model also allows for digital simulations, performance evaluation was not part of this particular design exercise. Finally, while all 3 processes are designed to be iterative, only MOCAPE actually uses an iterative method to refine the form based on structural performance.
As highlighted by a matrix comparing methods across projects and design phases (Figure 34), the case studies presented here show a varying combination of manual and algorithmic-driven processes. In each case, a unique design work flow reflects careful examination of the specific problem at hand. Interest-ingly, any remaining manual operation within these processes can be automated by using an algorithm, yet none of the projects relies exclusively on computation for generating form. In all cases, the formal solution does not result from any individual step or algorithm, yet it is never preconceived or induced. Rather, form emerges from a mediated process that augment human judgement with computational power.
The projects presented here reveal the same interest for a rich, intricate, and ultimately seductive formal complexity that informs some of the most prominent architectures being built today. However, formal complexity here reflects the complexity of the context at hand, rather than an arbitrary shape conceived by the architect. Each of the 3 formal solutions works with—not against—natural forces, resulting in economic and intelligent use of material and technology. In this context, a machine is better equipped to reproduce the complex logics and infinite formal intricacy we find in nature, but only if the designer is able to identify a suitable scope for it. Far from perfect, these heuristic methods engage in authentic dialogue with the natural environment in pursuit of a truthful relationship between energy and form, based on explicit rules and measurable results.
REFERENCES
BIM Software for Architectural Design | Autodesk. (n.d.). Retrieved October 06, 2017, from https:// www.autodesk.com/solutions/bim/architecture
Diamond, J. (2005). Collapse: How societies choose to fail or succeed. New York, NY: Penguin Books. Dutton, D. (2010). The art instinct: Beauty, pleasure, & human evolution. New York, NY: Bloomsbury Press.
Fuller, B. R., & In Snyder, J. (2010). Ideas and integrities: A spontaneous autobiographical disclosure. Baden, Switzerland: Lars Müller Publishers.
Giedion, S. (1967). Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press.
McCullough, M. (2013). Ambient Commons: Attention in the Age of Embodied Information. Cambridge, MA: MIT Press.
Progress on energy efficiency in Europe. (2016, December 6). Retrieved October 06, 2017, from https:// www.eea.europa.eu/data-and-maps/indicators/progress-on-energy-efficiency-in-europe-2/assessment-2 R. (1984). The architecture of the well-tempered environment. Chicago, IL: University of Chicago Press.
