B. Kolarevic-Performative Architecture

273

Transcript of B. Kolarevic-Performative Architecture

Page 1: B. Kolarevic-Performative Architecture
Page 2: B. Kolarevic-Performative Architecture

EDITED BY BRANKO KOLAREVIC & ALI M. MALKAWI

ARCHITECTUREBEYOND INSTRUMENTALITY

PERFORMATIVE

Page 3: B. Kolarevic-Performative Architecture

First published 2005by Spon Press29 West 35th Street, New York, NY 10001

Simultaneously published in the UKby Spon Press11 New Fetter Lane, London EC4P 4EE

Spon Press is an imprint of the Taylor & Francis Group

© 2005 Branko Kolarevic and Ali Malkawi, selection and editorialmatter; individual chapters, the contributors

All rights reserved. No part of this book may be reprinted or reproducedor utilised in any form or by any electronic, mechanical, or other means,now known or hereafter invented, including photocopying and recording,or in any information storage or retrieval system, without permission inwriting from the publishers.

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

Kolarevic, Branko, 1963- Performative architecture : beyond instrumentality / Branko Kolarevicand Ali Malkawi.

Includes bibliographical references and index. ISBN 0-415-70083-3 (hardcover : alk. paper) ISBN 0-203-53685-1 (ebook)1. Buildings--Performance. 2. Buildings--Energy conservation.3. Architecture--Decision making. I. Malkawi, Ali. II. Title.TH453.K65 2004720--dc22

2004014235

This edition published in the Taylor & Francis e-Library, 2004.

"To purchase your own copy of this or any of Taylor & Francis or Routledge'scollection of thousands of eBooks please go to www.eBookstore.tandf.co.uk."

ISBN 0-203-01782-X Master e-book ISBN

Page 4: B. Kolarevic-Performative Architecture

CONTENTS

ACKNOWLEDGMENTS ................................................................................. v

PROLOGUE (Kolarevic) ................................................................................. 1

1 ARCHITECTURE'S UNSCRIPTED PERFORMANCE (Leatherbarrow) ......... 5

2 PRODUCT AND PROCESS:

PERFORMANCE-BASED ARCHITECTURE (Whalley) ............................... 21

3 SUSTAINABLE DESIGN: AN AMERICAN PERSPECTIVE (Raman) ......... 41

4 BIOTECHNIQUES: REMARKS ON THE

INTENSITY OF CONDITIONING (Braham) ................................................ 55

5 PERFORMANCE FORM (Herzog) ............................................................... 71

6 PERFORMANCE SIMULATION: RESEARCH AND TOOLS (Malkawi) ...... 85

7 A FRAMEWORK FOR RATIONAL

BUILDING PERFORMANCE DIALOGUES (Augenbroe) ............................. 97

8 ENGINEERING COMPLEXITY:

PERFORMANCE-BASED DESIGN IN USE (Schwitter) ........................... 111

9 ENGINEERING IN A PERFORMATIVE ENVIRONMENT (Blassel) ......... 123

10 NON-STANDARD STRUCTURAL DESIGN FOR

NON-STANDARD ARCHITECTURE (Kloft) .............................................. 135

11 COMMUNICATIVE DISPLAY SKIN FOR BUILDINGS:

BIX AT THE KUNSTHAUS GRAZ (Edler) ................................................. 149

12 THE STRUCTURE OF VAGUENESS (Spuybroek) .................................... 161

13 PERFORMATIVITY: BEYOND EFFICIENCY AND

OPTIMIZATION IN ARCHITECTURE (Rahim) ......................................... 177

14 COMPUTING THE PERFORMATIVE (Kolarevic) ..................................... 193

15 TOWARDS THE PERFORMATIVE IN ARCHITECTURE (Kolarevic) ....... 203

16 PERFORMANCE (AND PERFORMERS):

IN SEARCH OF DIRECTION (AND A DIRECTOR) (McCleary) ................ 215

17 CONCEPTUAL PERFORMATIVITY (panel discussion) .............................. 225

18 OPERATIVE PERFORMATIVITY (panel discussion) .................................. 237

EPILOGUE (Malkawi) ............................................................................... 247

APPENDIX ................................................................................................. 251

AUTHORS’ BIOGRAPHIES ....................................................................... 253

IMAGE CREDITS ...................................................................................... 261

INDEX ....................................................................................................... 263

Page 5: B. Kolarevic-Performative Architecture
Page 6: B. Kolarevic-Performative Architecture

ACKNOWLEDGMENTS

v

We are indebted to the contributors of this book forthe time and effort they have devoted to writing thechapters and for putting up with the various editorialdemands.

We are grateful to the sponsors of the symposium“Performative Architecture,” which was held at theUniversity of Pennsylvania (Penn) in October 2003,and which led to this book. The symposium wasorganized by the Digital Design Research Lab (DDRL)and the Building Simulation Group (BSG) in theSchool of Design at Penn (PennDesign). It wassponsored by McGraw-Hill Construction, TurnerConstruction, Autodesk, Bentley Systems, Flomerics,

the Department of Facilities and Real Estate Services,the Department of Architecture at Penn and byPennDesign.

We would like to thank the University ofPennsylvania and the School of Design for providing anexceptional environment for teaching and research. Weare particularly grateful to Professor Gary Hack, Dean ofPennDesign, Professor Detlef Mertins, Chair of theDepartment of Architecture, and Professor RichardWesley, former Head of the Department of Architecture,for their support of our endeavors. Finally, we want tothank our students and our colleagues for the stimulatingconversations about the subject of this book.

Page 7: B. Kolarevic-Performative Architecture
Page 8: B. Kolarevic-Performative Architecture

1

BRANKO KOLAREVIC

PROLOGUE

Page 9: B. Kolarevic-Performative Architecture

2

BRANKO KOLAREVIC

PROLOGUE

Page 10: B. Kolarevic-Performative Architecture

3

This book discusses an emerging approach toarchitecture in which building performance is aguiding design principle. This architecture placesbroadly defined performance above, or on a parwith, form-making; it utilizes digital technologies ofquantitative and qualitative performance-basedsimulation to offer a comprehensive new approachto the design of the built environment.

In this new information- and simulation-drivendesign context, the paradigm of performance-baseddesign can be approached very broadly — itsmeaning spans multiple realms, from spatial, socialand cultural to purely technical (structural,thermal, acoustical, etc.). The increasing emphasison building performance — from the cultural andsocial context to building physics — is influencingbuilding design, its processes and practices, byblurring the distinctions between geometry andanalysis, between appearance and performance. Byintegrating the design and analysis of buildingsaround digital technologies of modeling andsimulation, the architect’s and engineer’s roles areincreasingly being integrated into a relativelyseamless digital collaborative enterprise from theearliest, conceptual stages of design.

The contents of this book emerged out of thesymposium on “Performative Architecture” held atthe University of Pennsylvania in October 2003.That event brought together some of the leadingindividuals from different realms — architects,engineers, theoreticians, technologists — who

comprise the contributors to this book, with the aim ofproviding informed views of what is meant byperformances in architecture and of architecture.

The idea for the symposium was based ondiscussions between Ali Malkawi and myself about theapparent “disconnect” between geometry and analysisin the currently available digital tools. Our initialscrutiny of the tools (the instruments) that are used todigitally simulate the performance of buildings — thetools that are increasingly accessible to architects andtheir consultants — provoked broader questions, suchas to what extent performance actually influencesdesign and what performance means in architecture.

As we engaged the theme in its broader dimensions,we discovered that little has been written aboutperformance in architecture. Yet this term —performance — has been widely used by owners,designers, engineers, cultural theorists, etc.Performance in architecture increasingly matters;however, it means different things to different people.

The chapters in this book provide a diverse set ofideas as to how building performance is relevant todayand how it might be relevant tomorrow for architecturaland engineering design practices. The projects discussedprovide snapshots of different approaches, grounded inactual practices already taking place.

As readers will notice, the meanings ofperformance in architecture are indeed multiple andintertwined, and are irreducible to a simple, succinctdefinition. Performance, however, will increasinglyunderlie discussions about architecture in the future.

Page 11: B. Kolarevic-Performative Architecture

4

Page 12: B. Kolarevic-Performative Architecture

5

1

DAVID LEATHERBARROW

ARCHITECTURE'S

PERFORMANCEUNSCRIPTED

Page 13: B. Kolarevic-Performative Architecture

6

1

DAVID LEATHERBARROW

ARCHITECTURE'S

PERFORMANCEUNSCRIPTED

Page 14: B. Kolarevic-Performative Architecture

7

The world is not an object such that I havein my possession the law of its making.

Merleau-Ponty1

I think that in every building, every street, there issomething that creates an event, and whatever creates anevent, is unintelligible.

Jean Baudrillard2

This chapter argues for a shift of orientation in architecturaltheory and practice, from what the building is to what it does,defining the first by means of the second. Broadly speaking,there are two ways designers and critics tend to view buildings:1) as objects that result from design and constructiontechniques; and 2) objects that represent various practices andideas. Although these accounts seem to explain fully thebuilding’s origin and destination, technological and aestheticstyles of thought reduce architecture to our concepts of it. Otherand essential aspects of buildings come into view if one supposesthat the actuality of the building consists largely in its acts, itsperformances.

The aim of this chapter is to outline how the buildingdiscloses itself through its operations (figure 1.1). For this to beapparent, constructive and perceptual intentionalities must betemporarily put out of play, not because they are misleading orwrong, but because they are normally taken to be fullyexplanatory. To see how the building itself operates,technological and aesthetic explanations must be temporarilysuspended. This means subordinating, at least for a while, thequestions about experience, meaning and production thatnormally occupy our attention. The autonomy thereby grantedthe building is contingent on this methodological premise. Atrisk in such an approach is architecture’s perfect rationality, forit will be seen that performances or events depend in part onconditions that cannot be rationalized. This does not mean theycannot be understood, just that they must be understooddifferently.

Before proceeding, a certain assumption aboutarchitectural performance needs to be rejected; namely, that thedevelopment of new instruments and methods of predicting thebuilding’s structural or environmental behavior will radicallyredefine the discipline’s practice and theory. Perhaps attentionto performance will contribute to a new understanding of theways buildings are imagined, made and experienced. But thisnew understanding will not result from the development anddeployment of new techniques alone. The continued dedicationto a technical interpretation of performance will lead to nothingmore than an uncritical reaffirmation of old-style functionalistthinking — a kind of thinking that is both reductive andinadequate because it recognizes only what it can predict. I willreturn to this point below.

1.1NeurosciencesInstitute, La Jolla,California (1992–95),architects Tod Williamsand Billie Tsien.

Page 15: B. Kolarevic-Performative Architecture

8

ARCHITECTURAL PERFORMANCEWhat is meant by architectural performance? Theterm is not new in architectural discourse, but itscurrent use draws upon non-architectural linguistictraditions. Is the performance envisaged for thebuilding like that of a machine or engine — the car onthe street or the stereo in the study — or, is it closer towhat might be seen on a theatrical stage or heard in aconcert hall? This inquiry’s central question can bestated simply: in what ways does the building act?What, in other words, does the architectural workactually do?

One way of resolving these questions rather quicklyis to say that the building acts to “house” activitiesand experiences; an auditorium, for example, houseslectures, likewise a kitchen cooking, a courtroom trials,and so on. While this answer contains a germ of truth,the life of buildings that is predicated on use can becharacterized as a borrowed existence, for it assumesthat the room’s or street’s recognizable profileconforms to our expectations of it.3 For the premise ofpredicated meaning to be sustained, the side of thesubject must be taken. The significance that buildingspossess is granted to them by you and me. Now, whatseems eminently sensible from a pragmatic andpedestrian point of view has been shown to be naïve inAldo Rossi’s critique of functionalism.4 Uses, he pointsout, often change throughout the life of the building —private houses become clinics, theaters apartmentblocks, and so on. A criterion so inconstant asfunctional use cannot, he suggests, be used to definethe building itself. Decisive instead, for Rossi, is type.For me it is operation or performance.

So a basic question presents itself: must the side ofthe subject be taken when an account of specificallyarchitectural modes of behavior is given? Might thisnot leave something out, perhaps something essential?Certainly buildings are designed and built “for us”: afarmhouse for farm life, a schoolhouse for schooling,and so on. A definition derived from Aristotle —architecture imitates human action and life — may beancient, but it is still largely true. Granting this, can

the building not also be understood apart from us and use,irrespective of programmatic requirements, individualdesires, and cultural expectations? If not fully, can it beunderstood at least in part, without turning to ourselves asthe benefactors of its identity? If we slacken the threads ofintention that bind us to objects, what will appear?

The question seems worth pursuing because it isundeniable that rooms predate our use of them. They alsoremain as they were once we have finished with them. Withjust this single and simple observation about the building’sextended temporality in mind, can it not be said thatarchitecture exists quite happily and completely without us,that it is not entirely determined by “anthropologicalpredicates” but is articulate on its own terms, that it is tosome degree un-predicated, even auto-predicated? The turnto “experience” in architectural discourse, often announcedwith all good intentions, is generally a secret turn to designand production, insofar as the perceived is taken to be whatis offered in designed perspectives. While congenial totechnical or professional interests, this turn might well causeus to miss the reality of the building itself — especially thatarchitectural reality that stands there irrespective of thevagaries of my interests or yours. My working hypothesis isthat the theme of performance is a key to the building’sinternal definition or pre-predicated existence.

1.2Philadelphia,Pennsylvania.

Page 16: B. Kolarevic-Performative Architecture

9

Again, there are two common ways of missing the reality ofthe architectural work: one is to see the building as nothingbut a system of components intended in design and realizedby construction, the other is to view it as a system ofrepresentations outlined in composition and experienced inperception. Both make the building into an object, the firsta result of technical reason and the second a confirmationof aesthetic expectations. This is precisely what happened inmodernist functionalism, especially in the version advancedin the post-World War II period, after modernism’s narrowdeterminism had revealed its incapacity for providing aplausible and legible urban architecture (figure 1.2). Thedebates on monumentality, for example, clearly testify to awidespread recognition of the poverty of functionalistsolutions in their call for new alliances with art practices inthe formation of urban centers. José Luis Sert’s theory5 andwork exemplify this stance very well, as they demonstratethe desire to couple technical and aesthetic concerns in theformation of new civic institutions. Marcel Breuer’swritings6 and buildings demonstrate a similar thesis. Mid-century writings that address technical problems also assertthe limited significance of functional concerns (the writingsof Olgay and Olgay7 are a good case in point, also the workof Max Fry and Jane Drew,8 for both pairs suggested thatart could compensate for the cultural sterility offunctionally determined solutions). The widely celebratedbuildings of our time no longer insert art into functionalsolutions, but they use it to drape or cover them: yet here,too, sculptural form is essentially a compensation for theinadequacy of functionalist solutions (figure 1.3).

Rather than rehearse this old debate between worksthat are useful and beautiful, seeking new answers toquestions that were poorly formulated in the first place, itmay be helpful to ask not about the work but about the waythe work works (figures 1.4 and 1.5). Is there “action” inarchitecture’s apparent passivity, in its steady and static

1.3Experience MusicProject, Seattle,Washington (1997–2000), architect FrankO. Gehry and Associates.

1.4The façade detail,Phoenix CentralLibrary, Phoenix,Arizona (1989–95),architect Will Bruderwith WendellBurnette.

1.5Phoenix CentralLibrary: The viewfrom within.

Page 17: B. Kolarevic-Performative Architecture

10

permanence? Is the application of the term “behavior”to architectural elements anything more than a patheticfallacy, or do buildings perform in some way? Comparedto dance and musical expression, the building seems tobe resolutely — even embarrassingly — inert andinactive; frozen music indeed. Compared to film,architecture seems positively motionless, about asanimated as a stop sign. The house, theater and museumjust sit where they have been planted, patiently awaitinga visitor’s arrival and experience, as if they could onlyglow with life when ignited by interests you and I bringinto them when we walk through their front doors. Butis the building only what we make of it? One suspectsthere must be something more to it because if it wereonly the consequence of an inhabitant’s intentions, itwould be impossible to understand why we often feel theneed to habituate ourselves to buildings, and also whythey can alternately depress and delight us.

EFFECTS, ACTIONS, AND EVENTSNo doubt the building is a technical and aesthetic work,but it is known as such through its workings, through itsinstruments and equipment (broadly conceived). Statedmore strongly, the building is its effects, and is knownprimarily through them, through its actions orperformances. What is true for people is also true forbuildings — character shows itself in what they do, inthe decisions, choices or actions they take. The reallocus and realization of character is in action — not insigns of identity added onto surfaces. The difficult pointis that these workings do not arise from capacities andconditions produced by technical reason or aestheticintentionality alone. One could assume these operationsintend the building to work “for us,” but, insofar as wewill be followed by others who will have differentexpectations, this must be a very generalized “us.” Thegenerality of this determination is so great that it isinsufficient to understand the building’s full actuality. Tograsp that, other kinds of performance must bedescribed — not performances in architecture, insteadperformances of architecture.

For the sake of an example, imagine a lecture theater.Certainly such a setting can be described objectively: thefour walls, the false ceiling that conceals air handlingequipment, the sloped floor, the seats arranged for theauditors, the lectern for the speaker, the gathering space onthe other side of the doors that benefits from naturallighting, and so on. All of it exists in indisputable factuality,all of it stands within the building on the site in the citywith impressive permanence and stability, waiting for eventswith unequaled patience. But is this availability, this“permanence in waiting” objective or object-like in the waywe commonly assume? Can such a room’s performance bemeasured like other objects in the world, will it yield itselfto techniques of sizing, weighing and computing? Is theroom a machine for living in?

The inadequacy of this conception can be seen if theroom’s performance over time is considered.

Consider first the room’s pre-history. Insofar as thelecture hall we have imagined has been arranged for theuses we intend, it existed before we walked though its doors.Because of this, it imposed itself on experience. Taking anabstract view of the situation, one can say some of itscharacteristics could have been foreseen. But in its concreteactuality every room is encountered as something donatedto us from a past into which we have no real insight, andover which we have absolutely no control. Of rooms wecommonly say they are given, but we generally have noknowledge of the people responsible for the gift, or of theirspecific intentions, desires and expectations for its use.Once seen, the room might be judged marvelous, singularlydepressing, or largely insignificant because it is so typical;but until it is seen in its concrete actuality, it is unknownand (in its specific qualities) unexpected. Whether great orsmall, there is always some surprise when buildings areentered, and this results from the fact that the particularityof each comes from a past of which we are largelyuninformed, each is “charged with a history that exceedsmemory,”9 the result of unknown and unknowableinitiatives.

In truth, we do not so much enter rooms, but rooms (soto speak) happen to us. One way to begin thinking about

Page 18: B. Kolarevic-Performative Architecture

11

what may be called the event-character of a setting isto consider its emergence out of a causality that noone understands very fully. Here we touch on an aspectof events that is essential — their unknowablebeginnings and unexpected occurrence. When we usethe expression “that was some event,” weacknowledge the unexpected quality of what occurred.We give such an experience the name event preciselybecause of the unforeseen character of what happened— real events are always more than what we expectedof them. Operations can indeed be managed, functionscan likewise be scripted, but the events we take asimportant cannot — or what is planned is not whatmakes them important. Similarly, events do not resultfrom technique or technical knowledge. This isbecause foresight is essential to technologicalthought.10 Technique is always anticipatory, it is aform of knowledge that leads to preconceived results.Because events arise out of a past that we do notknow, they cannot be produced technically. Putting thematter more forcefully, performative architecture isnot the outcome of building or design technology, evenup-to-the-minute digital technology. All that techniquecan give architecture is enhanced functionality.

The room’s “eventmental”11 nature is especiallyclear when its present appearance is considered.Viewing the way an auditorium presents itself on agiven occasion, one could, speaking in generalities orbroad abstractions, imagine the ways it subsists in its“indifferent emptiness” between various events. If theroom’s actuality is important, however, a wiserprocedure would be to observe and describe itsparticularity on a given afternoon, filled with aparticular event, on an announced topic, and so on.The setting stages an event, the success of which atany given moment is unknown. Is the room working?Are its provisions serving the aims of the speakers andaudience? One could certainly say the room isgenerally adequate, but not until the event unfolds —in its unparalleled particularity — will anyone be ableto say whether or not the room and the event was a

success or failure. Both the singularity of the occasion —the working of the room for a specific occasion — and its“undecidability” in the present must be stressed. Even ifinheritances confer orientation, they offer no guarantees.As with its past, the room’s present condition isunknowable, but also unrepeatable, and cannot beconstituted as such by any explicit intention. The event, aswe say, is or is not happening. No single contribution, nomatter how well-planned or thought out, can control itsunfolding, not the schedule of the presentations, therefreshments for the breaks, the equipment provided, northe soundproofing of the room will allow one to fullyanticipate the outcome. If the event is only what wasanticipated, it will have been both uneventful andunmemorable. Viewing the way an event unfolds in thepresent, we can discern an essential aspect of settings:place-bound events that truly merit the name arise out ofthemselves, despite my interests or yours, as if they wereindifferent to them.

Lastly, in this little sketch of the auditorium’stemporality, its future requires attention. How will a givenevent play itself out, a conference let’s say? How will theroom perform for a later speaker in the afternoon when thetime after lunch brings a little sluggishness? Can one say alistener’s inattention during a later speaker’s talk will haveresulted from the quality of the argument, the heaviness oflunch, or the way the room’s atmosphere blankets aperson’s awareness? And what about one’s understandingof the talks themselves, the differences between what thespeakers will want to say and what each of us will hear,about all the disagreements, distortions, silentunderstandings, exaggerations and so on? Can any of thatbe known objectively? In order to provide an adequateaccount of what will have occurred one would need anunending hermeneutic. The event in, and of, a particularroom is a phenomenon without a clearly known orknowable boundary, end or identity. What objects possessin abundance — sharp definition — events lack almostentirely. Put differently, the performance of a setting canonly be known on its own terms, or, as I suggested earlier,pre-predicatively. Events cannot be defined, organized or

Page 19: B. Kolarevic-Performative Architecture

12

scripted because their beginning, middle and end resistobjective comprehension. This leads to a first conclusion: tounderstand architecture’s performative character we cannotrely on transparent and objective description alone, or ontechniques of quantification and measurement.

THE DEVICE PARADIGMIn what types of places do architectural operations unfold?Where are architecture’s unscripted performances typicallystaged? One obvious answer is the building’s moving or (moreexactly) moveable mechanisms. This is an undeniable kind of“action” in architecture, for some of the building’s partsliterally move or allow themselves to be moved: apertures,screens, furnishings, etc., each of which has its own “rangeof motion,” its stops, levels, intervals, etc. that anticipateand regulate its shifts and repositionings. Movements of thiskind are variously manual and mechanical, initiated byhuman or environmental prompts, and controlled by manual,electrical or digital mechanisms. Their tasks, in general, arethe modification and mediation of the environment in itswidest sense, from climate to human behavior. The work ofRenzo Piano’s Aurora Place in Sydney, Australia (1996–2000), for example, could be used to illustrate this aspect ofarchitectural performance. Each of its exterior surfaces, andall their elements, consist of moving and moveablemechanisms (figure 1.6). Perhaps the most famous earlytwentieth-century example of a building that presents itselfas an ensemble of adjustable equipment is the Maisond’Alsace (Paris, 1928–31) by Pierre Chareau, with itsexceedingly elaborate apparatus of ladders, screens, shadesand so on. Roughly contemporary, and offering equivalentdevices on the building’s exterior, is Giuseppe Terragni’s Casadel Fascio (Como, Italy, 1932–36). Perhaps the grandestexample from these same years is Le Corbusier’s Palace ofthe Soviets (1931) — huge sections of the building weremeant to move with the actions of the assembled multitude.

Design of this sort follows what might be called thedevice paradigm.12 The positions each element can take —the stops, levels and intervals — script the device’sperformance. Typically, these positions outline or frame arange of movements, normally from open to closed (figure1.7). The intelligence of a device is measured not by the

1.6Aurora Place, Sydney,Australia (1996–2000),architect Renzo PianoBuilding Workshop.

1.7L’Institut du MondeArabe, Paris, France(1981–87), architectAteliers Jean Nouvel.

Page 20: B. Kolarevic-Performative Architecture

13

breadth of this range, nor by the number of intermediatepositions, but by its capacity to adjust itself to foreseenand unforeseen conditions. An analogy that may be usefulhere is with musical or theatrical improvisation, as if thestops and positions of the building’s elements do nothingmore than sketch out the guidelines of a performance,allowing for spontaneous qualifications that attune theensemble to particular conditions, as they vary over time.Approximate movements can be intended, but settings canalso yield, respond or react to unforeseen events. Thearchitectural drama, then, comes alive through thebuilding’s performances. The first step in the developmentof a performative architecture is to outline strategies ofadjustment.

ECONOMY OF PERFORMANCEThere is another site of architectural action in whichperformance is less obvious but no less determining: thoseparts of the building that give it its apparently staticequilibrium, its structural, thermal, material stability.When discussing these elements (columns and beams,retaining walls and foundations, but also cladding androofing systems), it is common to talk of their “behavior”— not only talk of it but to anticipate it, even predict it.Obviously, talk of this sort is metaphorical, but in truth thebuilding must work at staying as it is. It must work withambient conditions, such as gravity, winds, sunlight and soon. It must also work against these forces. And it mustsuffer their effects. No actor on stage ever suffered asmuch as buildings do — whether one thinks of use andmisuse, weathering, or additions and alterations.

The economy of performance — in a site, as if on astage — is always an exchange between forces andcounterforces. To act is to counteract. The building’s laboris quite simply the amount of effort it takes to sustain thiseconomy, to keep up or play its part. The term we use mostfrequently for this work is resistance. The work of TodWilliams and Billie Tsien, especially the NeurosciencesInstitute in La Jolla, California (figures 1.1, 1.8 and 1.9;1992–1995), from a few years ago demonstratesawareness of this kind of performance — working withand against its site — also, much more locally, against the

1.8The upper deck,Neurosciences Institute,La Jolla, California(1992–95), architects TodWilliams and Billie Tsien.

1.9Neurosciences Institute:the central court.

Page 21: B. Kolarevic-Performative Architecture

14

pressures of human touch, evident in its variously rough andsmooth surfaces. A similar, and well-known, case from thetwentieth century is Mies van der Rohe’s Tugendhat House(Brno, Czechoslovakia, 1928–30). Less well-known, buteven more vividly engaged with its site, is his Wolf House(Guben, Germany, 1925–26).

Speaking very generally again, the work the buildingperforms involves resistance, by means of which itscapacities and identity become apparent. The façade, in fact,is the site of precisely this resistance, offered by the latentqualities of materials against ambient forces (figure 1.10).Should one say, as I did earlier, that the building’s destiny isto suffer? Is its work passive? That depends entirely on whatis meant by the term. Is it fair to say that the sprinter poisedfor the start of a race is passive? Is it not more accurate toobserve that the explosion ignited by the starter’s pistolpresupposes a coiled potential that can only be constitutedand maintained through strenuous effort? Is the building’saction against the steady pressure of the hillside into whichit is cut any different? In both there is force andcounterforce, which suggests an inevitable contextuality ofthe building’s performative elements, by which its equipmenttranscends itself into a range of spaces and regions in thesame way that it transcends itself into several temporalities— disavowing, again, its status as an object or phenomenonthat can be objectively defined.

The design for performance of this sort is based not on adevice but on a topography paradigm. Movement here is notthe change of position but of state. The force–counterforcerelationship results in alterations to the building’s physicalbody that demonstrate its capacity to respond to ambientconditions. Stains on the building are evidence of itscapacity for resistance. Cracks in the wall show limitedsuccess on this front.

The obverse of cracking and staining is shaping andfinishing — whether that of construction technology,environmental influence or everyday use. This kind ofsuffering (architectural pathos) was described by PeterZumthor as enrichment.13 Buildings, he said, take on abeautiful and specific richness when traces of life aresedimented onto their surfaces (figure 1.11). Movement,action or performance in the so-called static permanence of

1.10Margaret EsherickHouse, Philadelphia,Pennsylvania (1959–61), architect Louis I.Kahn.

1.11Ise shrine, Ise,Japan, last rituallyreconstructed in1993.

Page 22: B. Kolarevic-Performative Architecture

15

buildings or elements is toward or away from the fullness oftheir potential — movement typically described asdevelopment or deterioration. Because not one of thesekinds of movement can be precisely predicted, architecturalperformance such as this can be described as unscripted.

Operations in and outside the building are dependent onseveral contingencies: those of the inhabitant’s interests andvantage, of the climate, the seasons, and of the times. Thebuilding’s workings are also dependent on changes to thoseparts that have been joined together to form the work. Nodoubt it is obvious to state that over time the building’smaterials eventually fail; but we rarely think seriouslyenough about this inevitable or essential indeterminacy.Materials suffer and vary at different rates. Some parts ofthe structure settle and move, others not, or not much,while still others suffer a range of surface variations. In theface of these alterations, maintaining equilibrium amongthe building’s parts is a task that cannot, in principle, becompleted. Nor can its difficulties be foreseen. In theunfolding of the operations that sustain a dwelling situation,architectural elements constitute themselves into somethingof a stable ensemble that possesses a comprehensible butprovisional finality, for such a configuration is always andonly temporary.

Concerning the changes the building suffers as a resultof “external” contingencies, there is some degree ofpredictability of developments, resulting from pastexperiences, but never certainty. Some locations showgreater constancy of climatic conditions, or less seasonalchange — such as the Caribbean — while others, likeCanada, show continual alteration. In Montreal, they say ifyou do not like the weather, wait ten minutes. But even inmoderate zones, the environment sometimes acts in ways itis not supposed to, as the history of tropical disastersproves. If the ambient environment was steadier in itsofferings, the building could assure itself of the adequacy ofits provisioning and would not need to continually adjustitself. In these circumstances (which are really those of thelaboratory), performances could be scripted. But the worldin which buildings actually exist is hardly so lawful.

The true measure of a building’s preparations is theircapacity to respond to both foreseen and unforeseen

1.12The Getty Center, LosAngeles, California(1984–97), architectRichard Meier andPartners.

Page 23: B. Kolarevic-Performative Architecture

16

developments. Stated in reverse, bad buildings arethose that cannot respond to unexpected conditionsbecause they have been so rigidly attuned toenvironmental norms. Our tendency to think ofenvironmental conditions as external or extrinsiccontingencies should be resisted, for no buildingoperates without them or apart from them. Jean Nouvelclaims to put nature to work in his architecture, for thebeauty and richness of his surfaces do not result fromdesign or construction technique alone but also fromthe action of ambient lighting, which variously andwonderfully saturates the skins with fluctuatingqualities — transparency, opacity, reflectivity and color.The environment — in this case, natural light — musttherefore be seen as internal to the building. Workingsof this sort are evidence of engagement between whatwas and what was not constructed, of the building’swillingness (or need) to interact with what it is not.Here again, non-objectivity, as contingency, enters theheart of the work, through its operations. Here, also,emerges the possibility of a form of representationinternal to architecture; for when the building identifiesitself with its milieu it becomes something it is not, theway all representational figures do. Put differently, thebuilding’s performance is the key to its natural(unimposed) symbolism, because when the buildingdefines itself in terms of what it is not (the natural andcultural milieu) it inaugurates precisely the sort of self-negation that is necessary for representation to occur(figure 1.12).

When the building is freed from technological andaesthetic intentionalities, we discover its lateralconnections to an environmental and social milieu thatis not of anyone’s making, still less of design andplanning. And it is precisely these connections thatanimate its performativity, even if they cause thebuilding’s work to resist both conceptual mastery andexhaustive description. The point to be stressed is thebuilding’s eccentricity, its existence outside of itself, forits behavior testifies to a constitutional weakness at itscenter, a negativity at its heart, because it must wait onthe environment to give it what it lacks — light, air,

human events and so on. Still, what the environment offersis always somewhat different from what was expected. Thebuilding’s internal disequilibrium obliges it to accept intoits make up conditions over which it has no control.

With the different dimensions of the building’scontingency in mind, a second conclusion can be proposed:that architecture’s performative labor has no end, for it isa task that continually presents itself anew.

TOPOGRAPHIES OF PERFORMANCEPerformance in architecture unfolds within a milieu that isnot of the building’s making. A name for this milieu istopography, indicating neither the built nor the un-builtworld, but both.14 Three characteristics of topographysustain the building’s performativity: its wide extensity, itsmosaic heterogeneity and its capacity to disclose previouslylatent potentials. There is always more to topography thanwhat might be viewed at any given moment. Excess isimplied in its ambience, for what constitutes the margins ofperceptual concentration always exceeds the expectationsof that focus. But this still more of topography, thisoutward increase of breadth and compass, does not offer toexperience more of what is locally apparent. Differencesare always discovered in the spread of topography; contrastand complementarity structure relationships between itsseveral situations and sites.

In modernist theory, space was presented as the all-embracing framework of every particular circumstance, theunlimited container of all possible contents. Likewise inmodern science, continuous space was understood to beisotropic and homogenous, possessing a self-samenesscongenial to intellectual mastery because of the conceptualcharacter of its attributes. The topography in whichbuildings perform is just the opposite of space: polytropic,heterogenous and concrete; its regions contrast, conflictand sometimes converse with one another. Yet it is not afield of infinite difference either, for it continually offersexperience of both unexpected and familiar situations. Ifspace advances its array all at once (in simultaneity),actual topography gives its locations through time. In anygiven site, at any given moment, its structure requires thatsome places be recalled, others anticipated.

Page 24: B. Kolarevic-Performative Architecture

17

Topography’s latency is apparent if one considers theway it gives itself to experience. Like events,landscapes — whether they are urban or not —contain unforeseen potentials, and show thesepotentials in the various ways they offer themselves toperception. The word “capacity” applied to physicalthings indicates similarly unseen possibilities.Capacities cannot be (fully) discerned because theykeep themselves recessive — like the backside of anobject one is looking at or the inside of somethingshaped or polished. Construction finishing aims tocultivate the potential of things. Because it too ismaterial, and can be cultivated, topography is notwhat physically appears in a given place — built orun-built — or not only that. Sites are surveyed in theearly stages of design so that given conditions can bedescribed and understood. While this seems obvious,the term given conditions is far from clear. We tend toassume that the place exhibits “its intentions” the waydesigns present theirs; in both, intentions are shown,and “givenness” we believe offers expressive display.But this again confuses the standing of a figure withthat of its ground, for the topography in whicharchitecture performs is not composed of objects inthe same sense; it does not expose the grounds(intentionality) of its formation, but serves as thegrounds for that formation (figure 1.13).

If topography’s potentials exceed one’s grasp and remainunforeseen to some degree, they can also be said to beunreasonable, at least in some measure. If, as arguedpreviously, the building is always or necessarily engagedwith topography, by virtue of its inevitable contextualityand contingency, its performances, too, will be (to somedegree) unplanned, or they will arise from “causes” thatare unassignable.

This suggests a different understanding of the building:it is not a technical preparation or not that chiefly, nor is itprimarily a representation of such a preparation, but it is anon-technical and non-aesthetic performance, thedesigner’s comprehension of which acknowledges itscontinual need for readjustments in order to reclaim itsown equilibrium and to sustain its engagement with un-built or previously built contingencies. Put more simply, thebuilding’s approximate disequilibrium animates a life anda history of ever-new performances (figure 1.14).

Aristotle once advised that the mark of a wiseindividual is to strive for the degree of exactitude indescriptions that is appropriate for the given subject. Thesame exactness, he said, must not be sought in alldepartments of philosophy alike, any more than in all theproducts of the arts and crafts. Let me cite him: “It is themark of an educated mind to expect that amount ofexactness in kind which the nature of the particular subjectadmits.”15 For this reason it is equally unreasonable to

1.13Santa Fe Art Institute,Santa Fe, New Mexico(1996–99), architectLegorreta + Legorreta.

Page 25: B. Kolarevic-Performative Architecture

18

accept merely probable conclusions from amathematician and to demand strict demonstrationsfrom an orator. In a similar vein Aristotlerecommended that when building a house, sketches ofbasic (configurational) principles should be made inoutline form only, so that they can be gradually filled inas unforeseen exigencies and opportunities arise. Thecarpenter and geometrician both seek after the rightangle, he said, but in different ways: “the former iscontent with an approximation to it which satisfies thepurpose of the work, the latter looks for the essence oressential attributes.”

INSTRUMENTALITY PLUSAt the outset I distinguished between two kinds ofunderstanding in the theory of architecturalperformance: the kind that can be exact and unfailingin its prediction of outcomes, and the kind thatanticipates what is likely, given the circumstantialcontingencies of built work. The first sort is technicaland productive, the second contextual and projective.There is no need to rank these two in a theory ofarchitectural performance; important instead isgrasping their reciprocity and their joint necessity. Ifacceptance of an uncertain foundation for performanceseems to plunge practice into irrationalism, we needonly remember that most of the decisions we make inour daily lives rest on a foundation that is just asuncertain. The cultural norms that serve as the horizonof unreflective existence will not stand up to rationalscrutiny, but are not for that reason nonsense, nor arethey opaque to understanding. They are certainlytransparent enough to sustain debate, the result ofwhich is adjustment or alteration. For a theory ofperformativity we should seek nothing more andnothing less: instrumental reason and the rationality onwhich it depends, plus situated understanding thatdiscovers in the particulars of a place, people andpurpose the unfounded conditions that actually prompt,animate and conclude a building’s performances.

1.14Carpenter Center for theVisual Arts, Cambridge,Massachusetts (1960–63), architect LeCorbusier.

Page 26: B. Kolarevic-Performative Architecture

19

NOTES1 Maurice Merleau-Ponty, “The Philosopher and His Shadow”in Signs, Evanston, IL: Northwestern University Press, 1964,p. 180.2 Jean Baudrillard and Jean Nouvel, The Singular Objects ofArchitecture, Minneapolis, MN: University of Minnesota Press,2002, p. 16.3 The term “borrowed existence” is derived from Jean-LucMarion’s In Excess: Studies of Saturated Phenomena, NewYork: Fordham University Press, 2002. While my argument isindebted to his entire account of “givenness,” his description ofborrowed existence can be found in chapter 2, “The Event orthe Happening Phenomenon.” I have also adapted hisinterpretation of the event structure of the lecture theater.4 Aldo Rossi, The Architecture of the City, Cambridge, MA:MIT Press, 1982, especially “Critique of NaïveFunctionalism,” and “How Urban Elements Become Defined.”5 See, for example, Sert’s contribution to the CIAM 8conference: “Centres of Communal Life,” in J. Tyrwhitt, J. L.Sert and E. N. Rogers (eds), CIAM 8 The Heart of the City:Towards the Humanisation of Urban Life, New York: Pellegriniand Cudahy, 1952, pp. 3–16. While his work is alsodocumented in this publication, see more fully Knud Bastlund,José Luis Sert Architecture, City Planning, Urban Design, NewYork: Frederick A. Praeger, 1967.6 See Peter Blake (ed.), Marcel Breuer: Sun and Shadow, ThePhilosophy of an Architect, New York: Dodd, Mead & Company,1956.

7 Most helpful among their writings on this particular topic is Aldarand Victor Olgay, Solar Control and Shading Devices, Princeton, NJ:Princeton University Press, 1957.8 See, for example, Maxwell Fry and Jane Drew, TropicalArchitecture in the Dry and Humid Zones, Malabar, FL: Robert E.Krieger, 1964.9 Op.cit. Marion, In Excess, p. 34.10 This point is elaborated in David Leatherbarrow and MohsenMostafavi, postscript to Surface Architecture, Cambridge, MA: MITPress, 2002.11 This term is Jean-Luc Marion’s. See In Excess, p. 32.12 This term, used more expansively, can be found in AlbertBorgmann, Technology and the Character of Contemporary Life,Chicago, IL: University of Chicago Press, 1984, especially seechapter 9.13 Peter Zumthor, “A Way of Looking at Things,” 1988, in ThinkingArchitecture, Baden, Switzerland: Lars Müller, 1998, p. 24.14 For more on topography in this sense see David Leatherbarrow,Uncommon Ground: Topography, Technology, and Architecture,Cambridge, MA: MIT Press, 2002, and David Leatherbarrow,Topographical Stories: Studies in Landscape and Architecture,Philadelphia, PA: University of Pennsylvania Press, 2004, especiallythe conclusion, “Ethics of the Dust,” which elaborates the pointssummarized here.15 Aristotle’s text, Nichomachean Ethics, 1.3.1, is cited anddiscussed in Wesley Trimpi, Muses of One Mind, Princeton: PrincetonUniversity Press, 1983, p. 125.

Page 27: B. Kolarevic-Performative Architecture

20

Page 28: B. Kolarevic-Performative Architecture

21

2

ANDREW WHALLEY

BASEDARCHITECTURE

PRODUCT AND

PERFORMANCE-PROCESS:

Page 29: B. Kolarevic-Performative Architecture

22

2

ANDREW WHALLEY

BASEDARCHITECTURE

PRODUCT AND

PERFORMANCE-PROCESS:

Page 30: B. Kolarevic-Performative Architecture

23

The evolution of late twentieth-century design can beinvestigated by considering some of its dominantthemes, such as mass production, informationtechnology, transportation and the workplace. It isalready evident that new concerns will influence thefuture of architectural and industrial design. If anytheme can characterize the new era, it is the changingperception of space and design’s new fluidity.

Architecture and industrial design are both aresponse to and reflection of the society that we livein. Architects produce designs by carefully analyzingrequirements and creating thoughtful solutions. Acombination of intellectual and manufacturingcapabilities enables them to do this — essentially, togenerate and develop the idea, the process and theproduct.

Grimshaw is a process-driven practice. There is nopreordained stylistic solution but rather a rigorousexploration of ideas resulting in a strong and legibleconcept — a design diagram. The solutions evolve outof a broad investigation and understanding of aproject’s program — the careful balancing of elementsthat make-up architecture. The investigation continuesdown to the finest detail, so that the overall concept isevident in the smallest part of the resulting building.

Grimshaw is also a product-driven practice thattakes a very pragmatic approach to architecture,which has evolved from an understanding ofmanufacturing and an appreciation of the way thingsgo together. This initially may appear to be an overtlyfunctional way of looking at things, but we believe thatout of that functionality, understood in performativeterms, beauty arises.

ARCHITECTURE OF CHANGETo understand the firm’s design philosophy and itsconsistent application over the past few decades, it isimportant to locate the practice’s work within thecontext of time and technological development.

The initial work at Grimshaw, in the 1970s, wasfounded in industrial architecture, of which an earlyexample is the factory in Bath, UK, for Herman

Miller, the furniture manufacturer (1976). The clientwanted a factory with the potential for change, because itdid not know what lines it would bring out over the nexttwenty or thirty years. The solution was an early use offiberglass paneling to provide a very flexible, adaptableskin (figure 2.1). Along with this adaptive cladding systemon the outside was a very flexible servicing strategy inside.The combination allowed the building to behave like anorganism that can adapt to suit different demands. Overfifteen years of use, the factory has been rearranged fivetimes. In the most recent change, the occupants moved thecanteen to an area that was formerly used formanufacturing and so it had an opaque skin. By movingglazed panels to that area, views onto the river wereopened up. By embedding adaptability into the design ofthe building and its systems, we created an architecturethat not only performs over time but that also improvesthe quality of life for its users.

For the second Herman Miller project at Chippenham(1982), also in the south west of England, Grimshaw againwanted a building that could anticipate change. It had tohave a large span and be essentially a big warehouse, butone with the scope for conversion to office use in thefuture. Long-term flexibility had to be anticipated. Wedesigned a whole cladding system (figure 2.2) that wouldmeet those performance criteria, because a warehouse isessentially about a skin. An office, however, is somethingmore sophisticated — you have to have windows, doorsand natural ventilation. The solution was to try to design askin that would foresee that change and that could beadapted to suit its use.

It soon became apparent that “off-the-shelf”components from cladding manufacturers were tooexpensive, or did not even exist. For the Herman Millerwarehouse, Grimshaw designed its own cladding system ofpressed aluminum — one which we have continued todevelop and improve over the years. To understand howelements such as cladding are made means going to thefactories and exploring the material and manufacturingpossibilities, which in turn informs the way architectsdetail and design a skin. Through this rigorous approachbeauty can come from pragmatism.

Page 31: B. Kolarevic-Performative Architecture

24

The practice pursued this approach in a number ofbuildings, such as the flexible factory designed for IGUS(Cologne, 1990–2000), where we have developed thevocabulary of the skin further with inspiration from JeanProuve. The physically expressed clamps show how the skinworks and how the panels are pressed back onto aframework (figure 2.3). That building is now up to phasefive and has grown over the last ten years to about 400%of its original size, which has been possible because it wasbuilt around courtyards and based on a design that didanticipate change. The system of roof lights brings lightdeep into the building. A series of modular elements floatinside the building, and can contain a variety of functions,such as office spaces or restrooms. The theater technologyof pressurized air is used in the pad units under modules,so each can be moved manually around the factory floor.Rather like a chessboard, one can rearrange the entirebuilding.1 The units can access natural light and ventilationno matter where they are placed, by simply hooking up toone of the roof lights.

In the Financial Times building (London, 1993), theconcept was all about the process of producing newspapersas part of a very dynamic, kinetic architecture. Commutersdrove past the building at night as the newspaper was

2.1Adaptive fiberglasspaneling system,Herman MillerFactory (1976),Bath, UK, architectGrimshaw.

2.2Adaptable claddingsystem, Herman MillerWarehouse (1982),Chippenham, UK,architect Grimshaw.

2.3The façade detail,IGUS factory,Cologne (1990–2000), architectGrimshaw.

Page 32: B. Kolarevic-Performative Architecture

25

being printed and then read it the next morning. Toshow the kinetic infrastructure, we designed a vastglass wall (figure 2.4), placing the structure on theoutside to articulate the building and to keep the insideunobstructed for the industrial process. But within tenyears of its opening, the building had a change of use.The printing presses went out and it became an Internetswitching center. Because of the robustness andflexibility in the architecture, the new owner made thechange in about six months. So to some extent, one canbuild in flexibility as part of a “loose fit” approach.

DESIGNING THE BUILDING’S DNAAt Grimshaw, a key concept, such as the large glasswall at the Financial Times building, is followed to thefinest detail. One can find the concept in the smallestdetail that makes up the building and understand thelarger picture from the finest elements. The sameamount of time and rigor are given to all the elements,however large or small (figure 2.5). In designing acomponent of a glazing joint, for example, thearchitects make prototypes in the office in foam andwood.2 These small elements can be seen as the DNA ofour buildings.

2.4The FinancialTimes PrintingPlant, London(1993), architectGrimshaw.

2.5Sketches and theprototype of the glazingbracket, The ChannelTunnel Terminal atWaterloo Station,London (1993),architect Grimshaw.

Page 33: B. Kolarevic-Performative Architecture

26

A very good example of that process, and one thatshows the shift to a more dynamic architecture, is theChannel Tunnel Terminal at Waterloo Station(London, 1993). The brief called for a very large roof,sheltering a terminal that had to contain many of theprogram items that would typically be found in anairport rather than in a railway station (figure 2.6).Designed to handle 15 million passengers, it also hadto be ‘shoe-horned’ into the center of the city. Therailway engineers produced a complex footprint toallow for the trains coming into the station. The brieffor the roof, which was only 10% of the capital cost ofthe whole project, had to enclose all that space,snaking its way to the terminus.

Again, a considerable amount of time was spentlooking at manufacturing as ideas were developed. We

made a series of models as a way of visualizing andunderstanding the space and structure, and then came upwith an efficient asymmetrical form to suit the tracklayout. It encompassed the best of both worlds. Whiledrawing inspiration from the great Victorian engineeredrailway sheds and triumphant halls of the nineteenthcentury, where the structure is expressed internally, it alsocreated an unprecedented public façade. The three-pin archform places the pin to one side, responding to theasymmetrical nature of the platform layout. The pin formsthe point of contraflecture, which means that thecompression and tensile elements reverse. We used this asa device to invert the relationship of interior to exteriorstructure. This exterior structure gave the building a publicface to present to London, something a Victorian stationnever did. Materials were only used where they were

2.6The Channel TunnelTerminal at WaterlooStation, London (1993),architect Grimshaw.

Page 34: B. Kolarevic-Performative Architecture

27

needed structurally. The use of telescoping andtapering tubes produced a dynamic skeleton-like form,with opaque cladding facing the station and a glasselevation towards city, with views of the trains comingin and out.

The problem was, of course, how to create a glassenvelope that could move and snake around theirregularly shaped site; for economy and expediency wewanted to keep all of the glass as a set of standardizedrectilinear sheets. So we designed individual elementsthat would create the whole (which goes back to theDNA concept mentioned previously). The key element isa joint shown in figure 2.7. The lattice has lots ofrectilinear sheets of glass similar to a Victoriangreenhouse. The joint allows the different geometries tobe used by letting each sheet slide, like the scales on a

snake’s skin. A considerable amount of time was spentdesigning a single component, a joint element, which couldpick up the skin anywhere in space. This is how the designconcept was fulfilled and the manufacturer satisfied. Theoverall effect is organic, fluid architecture, although all ofthe roof glazing is made out of rectilinear pieces of glassand tubular elements.3

While Waterloo was still being designed, Grimshawwon a competition to design the British Pavilion at theWorld Expo ’92 in Seville (1992). The aim was to createan environment suited to an Expo in a very hot climatewhile also demonstrating a high degree of sustainability.We designed an enclosure that tempered and controlled theenvironment (figure 2.8), and we placed within it a numberof highly conditioned pavilions to provide flexibility of usebecause the exhibition content was still to be established.

2.7Waterloo Station:the glazing detail.

2.8British Pavilion at theWorld Expo ’92,Seville (1992),architect Grimshaw.

Page 35: B. Kolarevic-Performative Architecture

28

The concept for this building was to use the sun to cool thebuilding. The roof was covered with photovoltaic cells and,together with fabric sails, shielded the enclosure from theheat. The resulting generated energy would pump water andpour it over the glass, allowing in light but keeping thebuilding cool. A reasonable thickness of water absorbsalmost all the infrared (heat) components of light, while stillallowing the rest of the visible spectrum into the building.This idea of functional performance was manipulated furtherto give the space exceptional qualities that warm or cool thesenses. Sculptor William Pye took the idea and developed theproject’s sculptural elements. He turned the water intodroplets, which visitors could hear falling down as theywalked through the space — they actually heard, saw andunderstood what was cooling the building. At night, thewater surface was lit and the visitors could still enjoy it forits intrinsic qualities. The idea of using water worked on anumber of levels — not just functional and not justperformance-related — because the space was also sensuousand sculptured.

DESIGNING THE CITY SPACESChange is also relevant on an urban scale. The Grimshawapproach can be seen in the Ludwig Erhard Haus (Berlin,1991–98), which contains the Berlin stock exchange andoffices (figure 2.9). The problem with office buildings in acity is that they impose their own grid and their own formonto the city’s streetscape. The main concept, and one whichis being followed in other new projects, was to create city-type spaces at ground level. The technically driven tectonicsolution was to hang the upper floors from a series ofstructural arches, an approach that opened up all sorts ofpossibilities. By hanging the floors, the maximum enclosureof office space could be created without infringing rights oflight and view angles. The resulting “soft” form is entirelydescribed by what the practice was allowed to build; realizingthis potential meant that it wasn’t necessary to design a tallbuilding. Consequently, the user can do anything at streetlevel because there is a clear spanning structure with nocolumns (figure 2.10). The elements that are in there nowcan be substituted with something else in the future, such asan internal plaza or a skating rink.

2.9Ludwig Erhard Haus,Berlin (1991–1998),architect Grimshaw.

2.10Ludwig Erhard Haus:Street level, interiorview.

Page 36: B. Kolarevic-Performative Architecture

29

Grimshaw also designed the elements that take the visitorsfrom that “city space” up into the offices. With thisbuilding, the brief was to create an office that would takethree or four different organizations in one space. In thelong term, it was not obvious how the building would beused, so it had to be organized to encourage people to mixand meet. The practice refers to this as the “mixing valveapproach” where the social aspect of work is opened up,which is essential, as it is the people using a space thatgive life to a building. A central staircase with broadplatforms encouraged everyone to change levels usingcentral circulation; people could also use a bridge in theheart of the building. The building allowed discourse andchance meeting, a social aspect of the office environment.

Similarly, Grimshaw’s work at Paddington Station(London, 1997–99) creates a city space where technologyhelps people use the space. Our aim was to reinvigoratethe Isambard Kingdom Brunel station of 1854, as it hadbecome very congested and we had to think of a new wayto make it work. A new building was built behind the mainconcourse (figure 2.11a), with new technological elementsto ease circulation (figure 2.11b). With the help of theGrimshaw industrial design team, banks of plasma screensin several locations disseminated information to stoppeople congregating around one centralized departureboard (figure 2.12). Introducing these performance-related elements within a 150-year old building instigateda new way of working, while respecting the historic fabricof the existing volume and space.

2.11a–bPaddington Station,London (1997–99),architect Grimshaw.

2.12Paddington Station:disseminatinginformation.

Page 37: B. Kolarevic-Performative Architecture

30

PERFORMANCE-BASED DESIGNOne of the best examples of change in buildings in theGrimshaw portfolio is the redevelopment of BatterseaPower Station (London, 2000–; figure 2.13). Designed tomake power out of burning coal, Battersea Power Stationis a massive, monolithic brick edifice that has been derelictfor over thirty years, and has been partly demolished.Grimshaw has designed a set of components to create alightweight roof for the old turbine halls. This solution —designed with the aid of non-linear mathematics — fusestwenty-first and early twentieth-century architecture tomake a vast and flexible enclosure for this mixed-usescheme.

Grimshaw’s brief for the Victoria & Albert MuseumBoilerhouse Extension (London, 1996; figure 2.14) was tocreate a twenty-first century enclosure for digital art. Wewanted to create a building that could showcase digital artand at the same time change, move, shift and replenishitself to reflect this ephemeral medium. A sculptureinspired the solution: a giant cube of ice by Anya Gallacio,in which the very beautiful platonic form slowlydisintegrated over time and morphed into something quitedifferent. Something crystalline and architectural, whichdisintegrates and changes, seemed to convey the idea ofdigital art and digital media. Again, technical solutionscame into play. At that time we used a new type of glass(Priva Lite by Saint Gobain) that could change from clearto translucent (and vice versa), and which could beprojected upon. When the glass is completely clear, onecould see the functions and exhibits within the building.When translucent, it could become a medium for showingtext, images or both. Like the ice sculpture, it could alsochange and seem to dematerialize.

We did not win the competition with that entry but, forGrimshaw, competitions are a chance to explore ideas in arelatively short space of time, as opposed to the often longgestation of a built project; in turn, these ideas can bedeveloped and incorporated into other projects if they arenot realized into buildings initially. In a successfulcompetition bid for a new art gallery in northern Spain,Fundacion Caixa Galicia (La Coruna, 1998–), which iscurrently under construction, we revisited some of the ideasfrom the previous project and took them forward. Thesurrounding buildings have beautiful “gallerias,”essentially double-skinned façades, which protect thebuildings from the harsh environment of the Atlantic. Afloating glass wall follows the line of these galleriafaçades, but the building then sweeps back physically to

2.13Redevelopment designfor Battersea PowerStation, London(2000–), architectGrimshaw.

2.14Victoria & AlbertMuseum BoilerhouseExtension, London(1996), architectGrimshaw.

Page 38: B. Kolarevic-Performative Architecture

31

allow light down into a very deep basement (figure 2.15).The building is almost as deep as it is high, and is shapedand sculptured by its performance criteria, but at the sametime it is sensitive to its setting. It was made using a newtype of glass (Holopro by G+B pronova GMBH), that has aholographic-etched interlayer which can be seen or whichcan disappear. If an image is projected onto it, from theoutside, it reads even in daylight as a sharp clear pictureon a solid surface, but from inside the image is not seenand one can look straight through it as with clear glass.With no projection, it appears as a clear glass surface fromboth inside and out, and from the street one can see thebuilding behind. So, by projecting onto the surface, theapparent position of the building envelope shifts and itsrelationship to the street changes — it is both a tectonicdevice and an urban mechanism.

At the Donald Danforth Plant Science Center (SaintLouis, 1998–2001), Grimshaw’s first project built in theUnited States, the brief called for a series of laboratories.These pre-described, highly serviced and functional spacesusually come in standardized layouts. We did not challengethis, but instead focused on how the building is set withinthe site. Given that Saint Louis is quite hot in summer andvery cold in winter, and that the circulation space wascritical, we created a central social space which connectedevery part. In our early schemes we explored the concept ofthe skin having two different relationships, depending onthe orientation. We designed a rigorously engineeredfaçade relating to the south-facing main street (figure2.16), protecting the building from the sunlight, but stilloffering limited views. From the other side, which isapproached on foot, the building presents a more organicfaçade with a louvered, sawtooth-like skin. The resultingeffect would have been that the building evokes a differentresponse from a distance and from close at hand. From theroad it is a precise, highly engineered building that one canlook into to see the central circulation space and thelibrary; one is given an entirely different experience whenwalking to the building from the adjacent parking lot.4

We have recently completed the initial design work fora new painting and drawing building for the Royal Collegeof Art (London, 2000–). Again, the design is performance-driven to create the optimum painting and drawingconditions. There are very few purpose built, high-qualitypainting schools. The most famous is probably CharlesMacintosh’s Art School in Glasgow, Scotland, which hasbeautiful north-facing painting studios. The existingpainting studios at the Royal College of Art receive direct

2.15Fundacion CaixaGalicia, La Coruna,Spain (1998–),architectGrimshaw.

2.16Donald DanforthPlant Science Center,Saint Louis (1998–2001), architectGrimshaw.

Page 39: B. Kolarevic-Performative Architecture

32

sunlight, forcing students to cover up the windows withpaper. The challenge was to work out how to createbeautiful light and space on a very tight site. The answerwas a skin that allows north light through but that shieldsthe building from direct sunlight. This approach hasresulted again in a building with a dual appearance: afairly opaque building on the side facing the Albert Hall (asolid building that sits comfortably with the monumentalbuildings around it) and a very open glass building as seenfrom Kensington Gore (which is due north) (figure 2.17).The zinc cladding gives the building solidity from the southside, but then, as one moves around the building, it has acompletely different appearance as it becomes more open.

At roof level, all the painting studios get the rightquality and balance of north light; a high-quality light isalso achieved for the studios on the lower levels (figure2.18). As already mentioned, the building’s design isperformance driven. The roof concept and the way theartificial and natural light are mixed required considerabletesting and the production of a series of different models.But, in addition to having a functional quality, there aresculptural and aesthetic qualities resulting from the waythe sun moves across a very deep, highly modeled façade;the sun dances across it, animating the louvers andcreating a very pleasing effect, which we simulated usingcomputer models.

Grimshaw’s work is evolving and this is partly due tothe way computers are changing how architects think aboutspace. An example is a small project for Cemex, a majorLatin American construction company. They centralized allof their facilities on one site in Mexico, commissioned anew computer that was going to run their entire globalorganization and called it HAL9000, in reference to thefilm 2001: A Space Odyssey.5 The HAL9000 project(Monterrey, 1996–97) is a simple concrete box that hasbeen manipulated to create a powerful identity for theclient (figure 2.19). The floor is actually a lighting system.The ceiling provides no light; instead, it is a fabric skinthat softly glows with reflected light. As well as being aninteresting visual experience, the solution is performancerelated in that the need for a raised computer floor allowedfor accommodating lighting at that level.

2.17The painting anddrawing building forthe Royal College ofArt, London (2000–),architect Grimshaw.

2.18Royal College of Art:partial section.

Page 40: B. Kolarevic-Performative Architecture

33

At Leicester University, UK, the practice was asked todesign a National Space Centre (Leicester, 1996–2001,figure 2.20) — an enclosure for rockets to excite andinspire the general public, and a base for the university’sspace research work. The performance objective was fairlysimple: creating a large enclosure for the rockets. Thechallenge was to design a fairly economic space whichremained light and airy. We decided to use a materialcalled ETFE foil, which is an air-inflated polymer that canbe used to glaze a structure with absolutely minimumweight. This helped to create a unique building, usingthree-dimensional form to create the architecture and to“consume” the geometries.

Our design for the Messehalle 3 trade fair hall projectin Germany (Frankfurt, 1999–2001; figure 2.21) usesgeometric solutions in a different way. The roof becomes apiece of origami — a folded plane. Our brief was simple: todesign a very large enclosure. The challenge was achievingthe 560-foot span. The performance criteria informed thesolution. The project team developed a single cord folded-plate roof to cover the span. It is this spanningfunctionality that gives it form and shape, both internallyand externally. Within that large hall many things couldhappen over the next 100 years — it is impossible topredict how it might be used. A flexible shape was created,but, again, rather than just designing a functional box, byresponding to the performance criteria of the brief, thearchitects gave the Messehalle a sculptural quality thatinforms the whole nature of the building.

2.19HAL9000, Monterrey(1996–97), architectGrimshaw.

2.20National SpaceCentre, LeicesterUniversity (1996–2001), architectGrimshaw.

2.21Exhibition Hall,Frankfurt,Germany (1999–2001), architectGrimshaw.

Page 41: B. Kolarevic-Performative Architecture

34

The next three projects show how computers help withthree-dimensional geometries, and the relationship betweendrawing and manufacturing. The first is the Eden Project(1996–2001; figure 2.22) in Cornwall, in south westEngland. The brief was simply to create large enclosures forplants — a twenty-first century botanical garden. One ofthe enclosures had to be big enough to allow a rain forest togrow to its full maturity, which meant creating somethingabout 150-feet high by about 300-feet wide in cross section(figure 2.23). That is considerably larger than traditionalglass houses.

The site was a quarry, and we thought that half thearchitecture was in the topography — the quarry’sincredible landscape. We captured elements of thatlandscape to create the enclosures; the architecture shouldhave the same fluidity and a synergy to the topography thatit would be nestled in. The initial ideas were similar to thoseof the Waterloo project — to use a series of nineteenth-century inspired trusses to create a sinuous linear structurethat would then be glazed.

Several factors led to a review of this approach, notleast the fact that the site was still being quarried and thetopography was changing throughout the design process.The radical solution was to create a series of “soapbubbles,” sitting lightly in the landscape. Each enclosurewas envisaged as a series of spheres made up of hexagonalpanels that could adapt to the changing ground plane byadding or subtracting panels as necessary. In practice, thismeant creating a computer model of the topography andone of the spheres (figure 2.24), and intersecting them todetermine the form of the enclosure.

Cutting through the model gave a changing fluid formas the topography of the ground and form of the enclosurechanged. The structure best suited to the resulting geometrywas made up of pentagons and hexagons. To minimize thesize of the structural sections, we used ETFE foil inflated

2.22The Eden Project,Cornwall, England(1996–2001),architect Grimshaw.

2.23The Eden Project:interior view.

2.24The Eden Project:the initial sphericalgeometry.

Page 42: B. Kolarevic-Performative Architecture

35

pillows to glaze the forms — we simply captured air with thestructure and the pneumatic skin (figure 2.25).

This structural geometry for the Eden Project was not anew idea. Not only was a model found dating from 1617, itis also prevalent in nature, as it is the most economical wayof capturing three-dimensional forms and creatinglightweight structures. The complication at Eden was that aseries of intersecting spheres was being used and the pointsof intersection gave rise to broken hexagons.

The Grimshaw project team decided to step back fromform and learn more from nature by looking at the way adragonfly’s wing works (figure 2.26). It is made up of panelsof skin that consist of hexagons and pentagons (figure 2.27).Where the panels meet, the hexagons break intoperpendicular lines, with a seemingly random pattern, butwhich actually follow the lines of force. The same processwas used for the spheres at Eden, allowing the structure tofollow the most efficient path rather than imposing somesort of grid form.

In performance terms, ETFE foil lets in more light thanglass, including ultraviolet light, so the plants inside canthrive better than they have ever done before in an artificialenvironment. It was also about a third of the price that aglass and steel enclosure would have been, meaning that theclient could afford it — an important consideration.

Most importantly, the enclosure needed to be as light aspossible for environmental reasons. If the enclosure wasmade out of small elements and from a lightweight material,it would require minimal transport to get the system down tothat remote part of England, which is poorly served by roadand rail. The whole weight of the roof, including the steeland the foil, is no heavier than the mass of the air itencloses. Ultimately, those are just facts and statistics, butoverall it is the Eden “biomes” as objects and theirfunctionality that gives them great beauty — the soapbubbles clinging to the rock face.

2.25The Eden Project:interior view.

2.26The dragonfly’s wing.

2.27The Eden Project:hexagons and pentagonsdefine the geometry ofthe enclosures.

Page 43: B. Kolarevic-Performative Architecture

36

DIGITALLY-DRIVEN DESIGN AND PRODUCTIONComputers allow architects and designers to follow the lawsof nature and to explore things in a much more three-dimensional way rather than being restrained by the meansof a set square, a drawing board and a standardmanufacturing technique. Computer systems are beingdeveloped to describe and define space; manufacturers nowown computer-aided design/computer-aided manufacturing(CAD/CAM) facilities to enable rapid prototyping, effectivelydoing away with notions of mass production. In the EdenProject, for example, that meant that each pillow could bedifferent without any prohibitive cost implications.

Other examples of the impact of CAD systems onarchitecture include a project that Grimshaw is designing inAustralia and a fourth phase at Eden. In Australia, thepractice is designing a whole city block that links theCentral Business District and the docklands (which arebeing redeveloped). Within that block, among otherbuildings, is the new Southern Cross Railway Station(Melbourne, 2002–; figure 2.28). The performancerequirements of the roof were a major consideration for therailway terminal. It acts as an umbrella or sunshade, but itneeded to be visually interesting as the skyscrapers aroundlook down on it. It also had to extract stale air from thediesel trains. The great fans that provide the obvioussolution to this problem, however, are neither sustainablenor aesthetically pleasing. Instead, we looked at theprevailing winds. The wind effectively sculpts and givesshape to the roof in the same way that it creates dunes insand or moguls in the snow. These dunes force the air topass over the roof surface, creating negative pressures to liftout and ventilate the space below. So the roof functionseffectively but is also visually interesting. It is theperformance criteria that give it shape and form — and itssculptural qualities.

On a much smaller scale, Grimshaw is now working onthe fourth phase at the Eden Project (Cornwall, 2003–).This phase comprises a series of buildings, including anotherbiome. Again, it required creating something different,pushing the performance criteria a stage further. One of thebuildings is an education center for which low-embodiedenergy construction is being investigated, as 1,000 schoolchildren on average visit Eden each week to learn aboutscience and biology; the client naturally wants the buildingto inspire them. The idea was that the building would tell thestory of transpiration, which is the way a tree uses energy.The resulting building is a tree-like shell form that follows alogarithmic geometry in the form of a spiral (figure 2.29).

2.28Southern CrossRailway Station,Melbourne, Australia(2002–), architectGrimshaw.

Page 44: B. Kolarevic-Performative Architecture

37

2.29Logarithmic, spiraling geometryfor the Education Centre, PhaseFour, the Eden Project,Cornwall, England (2003–),architect Grimshaw.

Page 45: B. Kolarevic-Performative Architecture

38

Although computers play their part, ideas originate inthe mind and are worked out with paper and pencil andsimple sketch models before they are transferred toCAD. It is only with the computer, however, that you canwork out the final geometry, the final connections andthe size of elements, etc. The practice had beendesigning components using rapid prototyping as a wayof quickly investigating the shape, performance andfunction. Now the architects can send a three-dimensional model to a rapid prototyping company anda day later get a model (figure 2.30). It is madeovernight, and it offers a great way to investigate formand shape. It can be worked on by hand, addingevidence, thinking about how the skin is going to work,the way it is going to allow the light in, and the way itis going to create a shelter. To that end, Grimshaw hasbeen doing a project with Bentley Systems looking atthe way the skin and structure can be described as aseries of changeable components and elements. The ideais that everything should be flexible and easily changed.

A new influence on the Grimshaw design process isthe practice’s Environmentally Viable Architecture(EVA) design guide — a software tool where architectscan measure a building’s potential impact on theenvironment. This encourages a performance-ledapproach to design and architecture, as the architect isgiven a continual feedback on the impact of their designand materials. The intention is to educate architectsand designers to be aware of environmental strategiesthat can assist in the production of sustainablearchitecture so that, ultimately, realizing a sustainablesolution becomes second nature. Critically, scores are ameasure of the process rather than the product.

PERFORMANCE-DRIVEN PROCESS AND PRODUCTArchitects are no longer constrained by the limits oftraditional construction techniques; designs can now befully conceived in three dimensions. More profoundly,architecture can be guided by the same laws that controland shape the world around us — an organic approach todesign based on exploring solutions through performance.

The practice is currently working on a project in NewYork State, a concert hall for the Rensselaer PolytechnicInstitute (Troy, 2001–; figure 2.31), developing interestingacoustic performance criteria by using fabrics in a new way.The real questions are: how can Grimshaw design anexperimental media center that, by its very nature, issomething that changes all the time? How do you meet thatchallenge architecturally? Who knows what is going tohappen in fifty years’ time? How can the architects expressthe potential for change in a functional way, in a“performative” way?

In this example, the answers came from physics. Themain concert hall is principally used for performingsymphonic work. It is a great acoustic space, and its shapeand form is entirely driven by creating the optimumacoustic space. This is the same performative criteria thatcreate a violin’s form. A Stradivarius violin is a functionalobject that produces the most beautiful range of sounds; itis this dedication to performance criteria that results in anobject that is also intrinsically beautiful to look at. Throughresponding to function logically, following the laws ofnature and physics, it is possible to achieve something ofexceptional quality. The Grimshaw project team wants theconcert hall to have permanence through the way itfunctions, by being shaped and formed through acoustics,like a musical instrument — it uses the law of physics butcan produce an object of great beauty in space.

Grimshaw’s work demonstrates that good architecturecan be process driven. Beauty can come from integrity, ifone responds to the functional requirements and designs atruly performance-related space. Following performativecriteria does not negate architectural qualities or beauty —when it is done properly, it generates them.

2.30The Eden Project,Phase Four, EducationCentre: a CAD/CAMmodel.

Page 46: B. Kolarevic-Performative Architecture

39

2.31Concert hall,Rensselaer PolytechnicInstitute, Troy, USA(2001–), architectGrimshaw.

Page 47: B. Kolarevic-Performative Architecture

40

NOTES1 It is not always easy to predict change. As DavidLeatherbarrow points out in Chapter 1, the user may alter theuse and the building still functions, but maybe in a way thearchitects never thought about. A practice can try to build intotheir building designs as much flexibility as possible, but it cannever completely anticipate the future.2 The early buildings were determined by what could bemanufactured and the way the practice thought and designed,using the set square and line. The advent of computers hasincreased the ability to meet targets, to describe, define andexplore spaces, and then ultimately to manufacture and makebuildings.3 This project is also a very early example of defining a formfully in three dimensions. It certainly could not have beendesigned without three-dimensional modeling software, whichwas used to describe and explore the complex forms in threedimensions.4 If this was a film then this particular scene ended up on theeditor’s floor for a series of reasons, but the overall buildingbudget clearly played a part. However, it was a theme that wecontinued to explore for the new project at the Royal College ofArt.5 Many of the things predicted in 2001: A Space Odyssey turnedout to be true. The film re-evaluates the idea of enclosure. Itchallenges surface and orientation — what is floor, ceiling, wall,what is volume. It offers a notion of very dynamic, changingenvironments.

Page 48: B. Kolarevic-Performative Architecture

41

3

MAHADEV RAMAN

PERSPECTIVE

SUSTAINABLE

AN AMERICANDESIGN:

Page 49: B. Kolarevic-Performative Architecture

42

3

MAHADEV RAMAN

PERSPECTIVE

SUSTAINABLE

AN AMERICANDESIGN:

Page 50: B. Kolarevic-Performative Architecture

43

The American experience with sustainable design differsfrom experiences in other parts of the world. When weunderstand the specific conditions that impact on theUnited States, we can explore whether sustainable designis needed here, where it should be practiced, who canpromote such designs and how we can measure theirsuccess.

Architects and engineers are exposed to well-knownenvironmental issues, including resource depletion,pollution and global warming. Professionals toutdefinitions of sustainability but the practical applicationof these concepts presents major challenges.Sustainability is not just about energy consumption — itis about finding the right balance between environmental,economic and social concerns.

What exactly are we trying to sustain? While muchdiscussion centers on saving the planet, the planet canactually take care of itself. If the human race waseliminated tomorrow, life on earth would continue toevolve and thrive. Therefore, sustainability is not aboutsaving the world — it is about saving ourselves andachieving as much longevity as possible for ourcivilization.

North America is the highest per capita consumer offuel in the world. The energy consumed by buildings in theUnited States represents approximately 40% of thisenergy consumption. Therefore, the building industry herehas an important role to play in shepherding this resource.Sustainability in buildings often means minimizing theconsumption of resources (water, energy, materials) butincreasingly it also entails maximizing the health, safetyand quality of life of their occupants. There are manychallenges to practicing sustainable design in ourenvironment.

This chapter explores the phenomenon of sustainabledesign from an American perspective and identifies theunique opportunities and obstacles designers face whenattempting to construct more environmentally responsiblebuildings in the United States. A number of case studiesillustrate the varying degrees of success achieved by thedesigners at Arup in pursuing a sustainable agenda onprojects completed in this country.

CHALLENGES TO PRACTICINGSUSTAINABLE DESIGNIn the disposable culture of the United States, there is asense that everything used can be discarded. Yet, whatever isdiscarded continues to exist in our one, closed system. Globalwarming is an example of the detrimental effect that ourlifestyles can have on the environment of a closed system. Inthe United States, where there are about 70 people persquare mile, we can continue to create and expand ourlandfills. That is not so easy in India and the United Kingdomwhere there are 700 people per square mile. Europe andChina have approximately 400 people per square mile. Theseenormous differences in population density breed variedapproaches to the environment. In the United States, we tendto think of our natural resources, including land, water andforests, as unlimited and disposable.

Legislative changes to codes and regulations do nothappen as quickly in the United States as in other countries.US citizens value their personal freedom greatly, as do theindividual states. Our federal government, for example,converted to the metric system several years ago but sinceprivate industry does not need to follow suit, the use of themetric system remains limited.

Code innovation occurs differently in the United Statesthan in Europe. In the United States, practitioners tend towork through the codes and regulatory structure as a meansof implementing change. Establishing a higher level ofperformance in buildings generally needs a consensus for theidea to build momentum so that eventually a code changefollows. In Europe, individual practitioners are alwayspushing the boundaries and this approach seems to besupported by the economics there. However, the downside ofthis approach is a higher variability in the quality ofbuildings.

Urban conditions present another challenge. Extensiveland resources in the United States have led to urban sprawl,resulting in increased traveling distances, traffic emissions,etc. This condition is less prevalent in more densely-populated areas with public transportation.

Another factor is the country’s climatic diversity. Thereis no single architectural style or engineering solution that isefficient and appropriate for buildings in all climatic regions.

Page 51: B. Kolarevic-Performative Architecture

44

Finally, financial issues shape our approach tosustainable design. Since energy prices for gasoline areapproximately 25% of prices in the United Kingdom, thepressure to conserve is not as great. These costconsiderations have a similar effect on powerconservation. Americans also demand much quickerpaybacks on investments than individuals in othercultures.

PROMOTING SUSTAINABLE DESIGNThe stakeholders in a sustainable design argument aremany. Owners bear the financial responsibility and haveconcerns over the long-range impact of a building’sdesign. Occupants have a stake but are rarely involved,so they can be seen as silent stakeholders. Governmententities and local authorities need to ensure that theircodes are not contravened. Designers and builders, ingeneral, wish to leave a positive legacy. The communityitself, both locally and globally, has to live with theconsequences of the design for many decades. Finally,there are the future generations of stakeholders, thepeople that we hope will, in perpetuity, be able to enjoythe quality of environment that we take for granted.Since the majority of these stakeholders are not engagedin the typical design process, the design team has a dutyto adopt the positions of others and to view the processfrom their standpoint.

Certain essential ingredients must be present in anyprocess that aims to achieve a better, more sustainableresult. There must be a commitment to the process,particularly from the owner, from the outset. A well-intentioned design team working without the owner’scommitment will not succeed.

There must be a non-traditional approach tocommunication and interaction among the disciplines. Atraditional sequential organizational structure will stiflethe best results. Communication that allows for feedbackand reworking is very important. A brilliant idea mayoccur spontaneously rather than in a scheduled meeting.A project must have an organizational structure thatwill allow for the capture of those ideas and the abilityto incorporate them into the process.

When a group of designers from different disciplines havecollaborated before, they often develop a rapport and a way ofplaying off each other’s ideas, which helps to hone the project.Today’s complex, modern building is the creation of manyprofessionals. The best examples of sustainable buildings areintegrated in such a way that the engineering is thearchitecture and the architecture is the engineering. It is theensemble that creates the desired results.

Also, recognition that various team members (architects,engineers, owners) communicate and process informationdifferently is key for success. The use of a variety of visualaids, animations, and simulations help in this process ofcommunication. Simple diagrams usually prove to be moreeffective than many pages of text in explaining energy flowsand other complex concepts (figure 3.1). Diagrams andvisualizations can also be especially effective whencommunicating across language barriers. When traditionalapproaches to presenting technical information through tablesand numbers are brought to life through animation, it becomeseasier to synthesize large amounts of data, not only for theteam members and code officials but for the presenter as well.

The final essential ingredient is a commitment to properlyfollow through. Typically, not enough is done to ensure that thedesign intent is being met in all of its depth. If there is not asubstantial commitment to follow through in the process, thenthe best design intentions will not be realized in practice.

MEASURING SUCCESSLimited budgets and fee structures are practical constraintsthat place a limit on the sustainable goals that can beachieved. Once stakeholders are identified, educated andcommitted, targets that seriously consider budgets and feestructures must be set. These targets can be very simple, suchas energy targets, and we can use grading systems such asLEED (Leadership in Energy and Environmental Design) tomeasure progress in meeting targets.

LEED is being adopted as a means of measuring abuilding’s sustainability. Does a LEED Platinum rating indicatethat a building is truly sustainable? Not necessarily. Rather, itindicates that the design team has achieved a certain amountof points on the LEED rating scheme. While LEED is anexcellent tool and provides some focus to a design team, the

Page 52: B. Kolarevic-Performative Architecture

45

system needs improvement and broadening. Setting targetsand securing points will not take the place of a gooddesign, but a good design team can take targets and turnthem into amazing results.

COST CONSIDERATIONSWhile first costs are often the foremost consideration inbuilding projects, operating and refurbishment costs needto be considered as well. How often do designers thinkabout the cost of removing and replacing a piece ofequipment in 25 years’ time? A smart design approach,which tries to address the varying lifespans of the differentbuilding elements and systems, can significantly facilitatefuture refurbishment.

Fees in the United States, particularly for engineers,are lower than elsewhere. But salaries for those sameprofessionals are higher. Consequently, the design processin the United States may use half to one-third of the man-hours devoted to the design process in the United Kingdom,for example. That is a staggering difference, whichseriously limits the amount of time that can be spent torefine designs and move them to a higher level.

The linear process is a natural response to the pressureto complete the design in fewer man-hours. In addition, thefact that schedules are tighter in the United States alsoleads to a more regimented process. To produce smart,sustainable designs within these constraints, designers needto be disciplined and focus efforts more at the earlyschematic design stage, and to test options while there arefewer constraints on the process. Once a smart schematicdesign is in place, the design needs to follow the morelinear process because it is not economically feasible towork otherwise. Innovation must be quick and early.

A sustainable design process, that is to be doneproperly, requires higher levels of professional fees than aconventional process. We can estimate that the schematicdesign may need 40% more effort and the designdevelopment perhaps 20% more. It may not require moreeffort to produce innovative construction documents butmore follow-through is imperative. Across a project, wecan estimate that 15–20% more fee is needed to deliver aninnovative sustainable design.

3.1aSunwall Sketch,winning entry for theDepartment of Energycompetition located attheir building inWashington, D.C.(2000), architectSolomon CordwellBuenz and Associates.

3.1bRendering of thedesign for theAmerican Universityin Cairo, Egypt (1999,not built), architectSasaki Associates.

Page 53: B. Kolarevic-Performative Architecture

46

CASE STUDIESOwner commitmentA client from the University of Texas proved hiscommitment to environmental issues by replacing aproject’s first architect when he felt that not enoughwas being done to incorporate environmental features.While the resulting building is not outwardlyremarkable to look at (figure 3.2), it offers a series ofinteresting features in terms of water recycling, solarharvesting, appropriate orientations to minimize solarheat gain, and superb use of daylight throughout. Thisbuilding is going to achieve a LEED Platinum ratingand the client is now pleased.

The Gap is committed to presenting anenvironmentally conscious and friendly image, andwanted to embody this philosophy in theirheadquarters building in San Bruno, California. Thesite selected by the Gap was inspiring because it sits ina climatic zone where buildings can be naturally

ventilated with relative ease. In the resulting building, air-conditioning loads are low and the design incorporates avariety of progressive features. Among the features is agrass roof that is designed to provide a certain amount ofphotosynthetic replacement for what has been taken awayfrom the site (figure 3.3). In the workspaces, no person ismore than 40 feet from a view and a source of daylight.Spaces like the atrium enjoy pragmatic approaches todaylight. The orientation of windows is vertical becausevertical windows are easier to seal. A curved sheetrocksurface efficiently reflects light down and into the officespace (figure 3.4). If the LEED system was in place whenthe Gap headquarters was constructed, the building wouldlikely have achieved at least a Gold rating. The Gap nowhas very high employee retention rates and this buildinghas probably already paid back any additional costs thatmay have been incurred by making sustainable designchoices.

3.2Rendering of theUniversity of TexasNursing School(currently underconstruction),architect BNIMArchitects.

Page 54: B. Kolarevic-Performative Architecture

47

Non-traditional approach to communicationand interaction among the disciplinesThe team organizational structure for a residence hall at theMassachusetts Institute of Technology (MIT) in Cambridge,USA (1999–2002, architects Steven Holl Architects; figure3.5) fostered a multidisciplinary, non-linear process thatrequired feedback. The owner community had four distinctconstituents, each with its own driving mission. The MITPresident desired a landmark building in terms ofarchitecture and sustainability; one goal was to develop thebuilding without traditional air conditioning. MIT’s projectmanagement team had a clear mission to deliver the projecton budget and on time. The facilities group needed assurancethat residents would be comfortable. Members of MIT’sBuilding Technology Faculty did not want to be associatedwith a naturally ventilated project because they did notbelieve the goal could be achieved. The design team includedtwo architectural firms, two structural firms, two Arupoffices and a host of other consultants.

3.3GAP Headquarters,San Bruno, California(1996–1998),architects WilliamMcDonough +Partners and Gensler.

3.4Interior of the Gapheadquarters.

Page 55: B. Kolarevic-Performative Architecture

48

While satisfying the differing interests of this group was aconsiderable challenge, the end project is a tremendoussuccess. In the plan that emerged, architecture, structureand building systems developed as one integrated whole.All building elements were designed to minimize thecooling load. The façade consists of 3,000 windows (figure3.5), each set deep within an 18-inch perforation thatallows for low winter sunlight to enter while shading therooms from the high summer sun. The thermal mass of thefaçade soaks up the solar heat, thereby minimizing thethermal plume that could enter the rooms via openwindows. This design enabled Arup to cool the building byway of a mixed-mode system that provides a very smallquantity of dehumidified cooled air whenever the outsidetemperature rises above 80ºF (figure 3.6). There have beenno complaints about occupants being too warm.

A project at New York’s Penn Station (1998–2003;with Skidmore Owings & Merrill) illustrates the success ofvisual aids. Designers studied the effect that light comingin through an expansive skylight would have on electronicdisplay boards being proposed for the station. If the

3.5Simmons Hall atMIT, Cambridge,USA (1999–2002),architect StevenHoll Architects.

3.6Simmons Hall:Atrium StudentLounge computationalfluid dynamics (CFD)study.

Page 56: B. Kolarevic-Performative Architecture

49

brightness of the light from the skylight could becontrolled without destroying the architecture of thespace, then designers could specify less bright and,therefore, less expensive display boards. Arup not onlyused a rendering package but also a photometricallyaccurate simulation to measure daylight, reflectionsand other criteria (figures 3.7 and 3.8). Designers thatare capable of such an advanced simulation canconduct these types of studies at the design stage,make decisions early, and move forward with a greaterlevel of confidence — an immensely valuable approach.Buildings can be constructed, examined and testedvirtually before a builder sets foot on the site.

At Norman Foster’s Greater London Assembly Building(1999–2001), a particular communication problem wassolved in an innovative way. The initial scheme for theassembly hall was a very smooth form. When the architectenvisioned the acoustics, he imagined that the sound of aperson’s voice would bounce up and escape out at the top ofthe shape. But acousticians explained that the assembly hallwould be excessively reverberant and that, without someabsorption measures, the acoustics would be terrible. Arupdeveloped a process to visualize the sound being reflectedand absorbed on different surfaces, and, after severaliterations, a solution emerged that was both architecturallyand acoustically acceptable (figure 3.9).

3.7Lighting simulation,Penn Station, NewYork, USA (1998–2003), architectSkidmore Owingsand Merrill.

3.8Lighting simulation designed toinvestigate the visual clarity ofelectronic displays for proposedPenn Station passageway.

3.9Acoustic progression,Greater LondonAssembly, London,UK (1999–2001),architect Foster andPartners.

Page 57: B. Kolarevic-Performative Architecture

50

Simulations can be applied to many design conundrums.One particularly challenging issue is designing escaperoutes in complex environments, such as, for example, largetransportation buildings with multiple levels. In theseunconventional buildings, it is often difficult to includemultiple stairways to meet the letter of the code. However,the openness of these buildings often means that occupantscan find safe escape routes with ease during an emergency.One powerful tool to demonstrate building safety is acomputer program called Simulex which enables designersto run an emergency evacuation simulation that considersdemographic data of building occupants, including age andagility. The program produces very specific real-timeanimations that demonstrate the adequacy of escapeprovisions and accurately estimate the amount of timenecessary for all occupants to escape from the building(figure 3.10).

Commitment to follow throughSite supervision, commissioning, handover to the facilitiesgroup and the management of a facility are the keys toachieving positive results. Arup’s client for Terminal 4 atJFK Airport in New York (1996–2001, with SkidmoreOwings and Merrill; figures 3.11 and 3.12) fully understoodthese keys; the same client was responsible for funding theproject, building it and operating it for an extended numberof years. The client paid the design team to maintainapproximately thirty on-site personnel throughout theconstruction to ensure that the design intent (including anadvanced integrated IT system) was fully realized. The teamwas able to immediately troubleshoot any obstacles toachieving performance goals during construction. It thenoversaw the painstakingly detailed commissioning of allsystems. In Terminal 4’s first year of operation, the client isreaping the benefits of its investment. While the building isapproximately twice the volume of the terminal it replaced,it uses only about 60% of the energy.

3.10Simulexevacuationpattern.

3.11Terminal 4 at JFKAirport, New York,USA (1996–2001),architect SkidmoreOwings and Merrill.

3.12JFK Terminal 4:linear skylightsprovide daylighting.

Page 58: B. Kolarevic-Performative Architecture

51

Measuring success and cost considerationsA major issue at Columbia University’s Lerner StudentCenter in New York (1994–99, with Bernard Tschumi;figures 3.13 and 3.14) was exposed steelwork. The steelworkwas critical to the architecture but the code required it to befireproofed. Arup conducted a fire engineering study thatillustrated there was no credible threat that fire loads couldcompromise the structure during the time it would takeoccupants to escape. A strong case was presented to thebuilding department and, after careful study and a period oftechnical arguments, the building department acceptedArup’s position. The study saved approximately $750,000 inintumescent paint costs; intumescent paint is notoriouslydetrimental to the environment.

3.13Lerner Student Center,Columbia University,New York, USA(1994–99), architectBernard Tschumi.

3.14Close-up of exposedsteel pedestrianwalkways at LernerStudent Center.

Page 59: B. Kolarevic-Performative Architecture

52

Richard Meier’s design for the Phoenix Federal Courthouse(1995–98; figure 3.15) involved the construction of a glassbox in the Arizona desert. Using computational fluiddynamics (CFD) and other thermal simulation techniques(figure 3.16), Arup devised an evaporative cooling systemthat conditions just the base of the atrium where the peopleare located. The solution cut approximately $5 million offthe cost of conditioning the space and reduced theoperating costs by approximately $50,000 annually. Thedesign, however, is not necessarily a paragon forsustainability. Arup took a space that would be intrinsicallyhot and would require enormous amounts of energy to aircondition, and instead introduced water as a means ofcreating an acceptable comfort level for the occupants.Normally, water would be considered a scarce and preciousresource in a desert climate. However, with long distancepipelines and massive pumping systems, our society is ableto supply millions of gallons of water to Phoenix at a very

3.15Federal Courthouse,Phoenix, USA (1995–98), architect RichardMeier and Partners.

3.16CFD thermal study ofthe Phoenix FederalCourthouse.

Page 60: B. Kolarevic-Performative Architecture

53

low cost. In this case, it is difficult to compare therelative environmental impact of reduced energyconsumption with that of increased water consumptionin a desert environment. In the LEED system, a buildingearns points for saving energy and water, but the LEEDsystem does not address environmental comparabilityand regional climatic issues. In the absence of concreteanswers and guidelines, designers need to ponder theseissues and use their best judgment.

Historical approaches to energy and technologyThe Chrysler Building (1930) was designed as anextension of the conventional punched window masonryfaçade building with no concern for air conditioning. TheEmpire State Building (1931) has a similar façade butintroduced early air-conditioning systems. A newapproach is evident in the United Nations Headquarters(1947; figure 3.17), with its clear single-glazed curtainwall. Here architects were “liberated” in their approachto design by the ability to install air conditioning. Aspart of a capital master plan, Arup conducted a surveyof the property and revealed the extreme care taken byboth designers and builders on the project. The UnitedNations Headquarters is three-eighths of an inch out of

vertical over the entire 36 stories and the stone’s condition isamazing. This is attributable in part to the architects’approach to the stone. The architects selected a stone that haspockets on its surface which hold moisture. To eliminateconcern over freeze-thaw damage, the façade is designed witha large heating coil on the back of the wall — an approachthat illustrates the disregard for resource conservation at thetime!

Part of Arup’s master plan brief was to study how toimprove the façade. Curtain wall panels were removed inseveral places where damage and corrosion were most likelyto occur. Surprisingly, the building had been built so well thatmost components were like new. Arup reported that, on thebasis of the condition of the curtain wall alone, there was nojustification for changing it. Arup then looked at the energyissues and found that the cost of replacing the clear single-glazed panes with high performance double-glazing wouldtake over 25 years to achieve a payback in terms of energysavings. This clearly shows that, while high performancefacades are environmentally sound, they are actually not costeffective to use in the United States, based on current energyprices.

CONCLUSIONThe future is bright for sustainable design in the UnitedStates. It will be even brighter when a clear connection canbe drawn between environmentally sound buildings and higheroccupant productivity.

People intuitively understand that being in a betterenvironment makes one feel better and more productive, but itcan be very hard to put numbers to it. While the connection isdifficult to measure, there are some recorded cases. RandolphCroxton designed a new day-lit factory for VeriFone adjacentto a conventional factory. The two buildings have the samekind of area and house similar activities. The old building is adumb opaque shed with industrial lighting while the newbuilding has a lot of daylight. The productivity in the day-litspace is much higher and people fall ill less. Apparently thereis a long waiting list of people to move from the old to thenew building. In this case, the productivity gain is clearlydemonstrated, and the economic benefit to the business canbe analyzed and documented.

3.17United NationsHeadquarters, NewYork (1947), with theChrysler Building inthe background.

Page 61: B. Kolarevic-Performative Architecture

54

As concrete evidence mounts and as people begin tounderstand the true value of environmentally sensitivedesign, opportunities to fund, design and construct well-integrated buildings will increase. It will becomepractical for owners and stakeholders to invest the timeand financial resources required, both for early-stageplanning and, critically, for on-site follow-through.

The current challenge, then, is to demonstrate thedirect relationship between the environmental quality ofa building and business profits. A 1% increase in staffproductivity can easily justify a 5–10% increase in theconstruction budget. Owners will come to understandthat in addition to the moral case for sustainable design,encouraging environmentally progressive design can alsobe a smart business move.

Page 62: B. Kolarevic-Performative Architecture

55

4

WILLIAM BRAHAM

CONDITIONING

BIOTECHNIQUES:

INTENSITY OFREMARKS ON THE

Page 63: B. Kolarevic-Performative Architecture

56

4

WILLIAM BRAHAM

CONDITIONING

BIOTECHNIQUES:

INTENSITY OFREMARKS ON THE

Page 64: B. Kolarevic-Performative Architecture

57

PERFORMANCE DESIGN (AGAIN)In 1967 Progressive Architecture magazine published aspecial issue on “performance design,” explaining it asa set of practices that had emerged from generalsystems theory, operations research and cyberneticsthirty years earlier, at the end of the World War II.1 Theeditors described its practitioners as “systems analysts,systems engineers, operations researchers” and arguedthat it was a more “scientific method of analyzingfunctional requirements,” which involved “psychologicaland aesthetic needs” as well as physical measures ofperformance. The interest in performance clearly drawson the long history of determinism and functionalism inarchitecture, understood in large part through themechanical and organic analogies of the late nineteenthand early twentieth centuries. It is perhaps fitting at theoutset to recall that Le Corbusier’s famous descriptionof a house as “machine for living” was his adaptation ofthe phrase that he and Ozenfant had earlier used todescribe painting, a machine a émouvoir — a machinefor moving emotions. All the objectivity of functionalmethods depends on the assessment of subjective needs,of quantified and temporarily stabilized desires.

To enter the discussion of performance design(again), this chapter examines the environmentalperformance of contemporary buildings. In the lasthalf-century, buildings have become bigger in a new andbulked-up sense; they enclose ever larger volumes,which have been engineered for ever greater comfortand productivity. This bulking-up of modernconstruction has been made possible by its systems ofconditioning — air conditioning, artificial illumination,

plumbing, electric power, telecommunications, and nownetworked information flow — which allow them toassume radically new scales and configurations.

To describe the mechanisms underlying theintensification of conditioning, I have adopted FrederickKiesler’s provocative term “biotechniques,” with all itsimplications of equivalence between biology andtechnology. In the current context, biotechniques mightbest be described as the biological analysis of technologicalsystems. They represent the collapse of the mechanical andorganic analogies in architecture within the powerfulconcept of complex system dynamics. The intensification ofconditioning operates equally on buildings and theirinhabitants, literally conditioning them to want and then“need” the new services, and steadily escalating the levelsof comfort and convenience they expect. That process hasits thresholds of intensity, beyond which results can be bothunexpected and difficult to reverse.

BIOTECHNIQUESThe term “biotechniques” was coined by the architectFrederick Kiesler in 1939 to indicate the equivalencebetween biology and technology.2 He was affiliated withBuckminster Fuller’s Structural Studies Associations atthe time and he used the term to distinguish his thinkingfrom the more direct imitation of biological forms orprocesses, which today we call biomimicry, and was beingcalled biotechnics by Patrick Geddes, Louis Mumford andKarel Honzik in Kiesler’s time.3 As he observed in anacerbic footnote, “[the Crystal Palace] was built by Paxtonin 1851 in imitation of the African water lily’s foliate, withits longitudinal and transverse girders. This was anessentially romantic attempt to fashion a man-builtstructure by literal application of nature’s designprinciples.”4 Instead, Kiesler based his term on a concepthe called “correalism”, by which he meant “the dynamicsof continual interaction between man and his natural andtechnological environments.”5 I do not mean to claimKiesler as the originator of these ideas — they were beingexplored in many fields — but he saw earlier than othershow radical their implications were for architecture. Thoseimplications derived from three basic propositions: first,that technology was based on steadily evolving humanneeds; second, that despite their origin in human needs,technological systems develop according to their own “lawsof heredity;” and third, that the final criteria oftechnological design is not technical performance, buthuman health (figure 4.1).

4.1“Man = Heredity +Environment. Thisdiagram expresses thecontinual interactionof both the totalenvironment on manand the continualinteraction of itsconstituent parts onone another.” FromKiesler, “OnCorrealism andBiotechnique.”6

Page 65: B. Kolarevic-Performative Architecture

58

In my adapted usage, biotechniques are any method bywhich buildings are examined as participants indynamic, “living” systems, whether of the biosphere orof financial, technical or social systems. They may ormay not produce results that look biological, and theywere initially deployed metaphorically to explain orunderstand how buildings or artifacts changed oradapted through time. Such biological analogies becamemore substantial with the introduction of devices andsystems that literally flowed or operated — plumbing,electricity, heating, ventilation and lighting — especiallywith the introduction of feedback techniques, like

thermostats, CO2 sensors and daylight monitors, that enabledbuilding systems to adapt and respond independently. Asthese elements were fixed in products, codes, standards andprocedures, the building of flows and its feedback devicesbecame the legal norm, while new techniques emerged tounderstand and regulate the dynamic aspects of design.

Such biotechniques became ever more important in thedecades after World War II, as cybernetics and generalsystems theory were applied to organisms and artificialsystems alike, rapidly collapsing the difference betweenmechanical and organic analogies, and making bothincreasingly operative. This is a critical point. At the momentthat living organisms (or ecologies) are understood as kindsof feedback systems, then the difference between mechanicaland organic systems virtually disappears. And almost fromthe beginning of systems research, natural and artificialsystems were analyzed together.7 The career of Jay Forrester,who developed the World III model used in The Limits toGrowth,8 exemplifies this process. After early work on airdefense systems, he focused his efforts on IndustrialDynamics,9 evaluating the dynamic problems inherent inindustrial production, sales and advertising, such as seasonalcycles, countercyclical policies, price stability, sensitivity andunexpected responses to all manner of events, actions anddecisions. Through a chance meeting with an ex-mayor ofBoston, he applied the same techniques to UrbanDynamics,10 and then after a conversation with the Club ofRome applied them to World Dynamics,11 exploring theinteraction between population, industrialization andpollution. This kind of world and climate modeling wascentral to the developing awareness of global environmentaleffects, making the construction and authorization of suchmodels vital (figures 4.2 and 4.3).

4.2Simulation model of“production-distribution system,”from Forrester,IndustrialDynamics.12

4.3Amplified oscillatoryresponse of sales andfactory inventory tothe introduction of afeedback mechanismin advertising, fromForrester, IndustrialDynamics.13

Page 66: B. Kolarevic-Performative Architecture

59

There have been many criticisms of these simple models,mostly that Forrester’s results exceeded the precision ofany data that were available. In defense, he argued thatthe “interaction between system components can bemore important than the components themselves” andthat the “computer model embodies a theory of systemstructure.”14 In World Dynamics, his primary interestwas global population and the early models capturedwere the dynamic, non-linear effects of multiplefeedback conditions; the effect of pollution, foodproduction and resource shortages on population andthen of population on food, pollution and resources. Butlike the contemporary simulations of artificial life, whatthese simulations lacked were any of the surprising andinnovative developments that seem to characterizeactual events, or even the internal “laws of heredity” oftechnological systems. They could not simulate theunpredictable effects that occur at certain intensities ofpopulation, such as occurred in the political transitionfrom city-state to national political organization or inthe technological transition from wood to coal, oil andgas.

The power of such models lies in theirdemonstration of effects that are complex, non-intuitiveor disproportionate to our actions. For example, manykinds of traffic jams occur once a certain number ofpeople decide to drive, once a certain threshold volume

of cars is on the highway. The creeping or stop-and-starttraffic that results is not caused by any one person’s speedor decision to drive, but occurs like the change of phase asa freely flowing liquid congeals into a solid at a certaintemperature (and pressure). One of the greatest challengesfor environmentalists is to demonstrate the connectionbetween seemingly minor individual actions — driving tothe supermarket, turning on an air conditioner — andthese kinds of threshold effects. And if the model is moreimportant than the data, the question for any dynamicsimulation is what flows and connections to model? AsForrester’s early work suggested, the critical sources ofenvironmental problems are ultimately social, cultural andpolitical, deriving from ideas about health, wealth andpleasure.

BIOTECHNIQUES: MORPHOGENETIC PRACTICESFor over a century, architects have sought “organic”techniques for generating building form, deriving themfrom structural diagrams, from charts of function, andnow from flows of data made manifest with digitalanimation software. The interest in these new techniques isnot difficult to assess. In a 1996 article entitled “Blobs (orWhy Tectonics is Square and Topology is Groovy),” GregLynn argued that “the mobile, multiple, and mutable body,while not a new concept, presents a paradigm of perpetualnovelty that is generative rather than reductive.”15 Thenovel morphogenetic properties of the new models aremade possible by the development and animation of“’isomorphic polysurfaces’ or what in the special-effectsand animation industry is referred to as ‘meta-clay,’ ‘meta-ball’ or ‘blob’ models.” Lynn explains that “in blobmodeling, objects are defined by monad-like primitiveswith internal forces of attraction and mass. Unlikeconventional geometric primitives such as a sphere, whichhas its own autonomous organization, a meta-ball isdefined in relation to other objects. Its center, surfacearea, mass, and organization are determined by otherfields of influence.” Those “fields of influence” can be usedto simulate anything from the motion of the sun to themovement of people to changing brand identities, anythingwhose influence can be assigned a value (figure 4. 4).

Critics like Michael Speaks have noted the apparentcontradiction between the responsive dynamism of theseanimate models and the inherently static nature ofbuildings.17 Speaks used his critique of novel andautonomous form to ask for a more flexible form ofpractice, in effect, opening design processes like that

4.4Dynamic,morphogeneticdesign model, GregLynn, Cardiff BayCompetition, 1994,from AnimateForm.16

Page 67: B. Kolarevic-Performative Architecture

60

described by Lynn to the fluid demands of the market.One could certainly point to many forms of architecturalpractice that have adapted quite aggressively to marketforces, from corporate design-build to the absorption ofprofessional designers into large companies. But thecritical aspect of these morphogenetic practices lies inthe use of explicit techniques to describe the flows,forces or elements influencing the production ofbuildings. For the most part those designers have wiselyavoided the fully deterministic conclusion of theirtechniques, using them as generative components inotherwise conventional design relationships. But a fewhistorical examples suggest just how challenging a fullydynamic account of design might be.

In 1940, the distinguished anthropologist A.L.Kroeber and a recent student of his published an articleon the “quantitative analysis” of “women dressfashions.”18 They charted the skirt and waist dimensionsof women’s fashions over three centuries, producingsome fascinating diagrams that showed the tendencies,trajectories and limits within those basic clothing forms.Certain limits are physical — dresses can only get sowide or narrow — while the basic trajectories appear tohave their own momentum, like Kiesler’s technological“laws of heredity.” Kroeber and Richardson were

careful not to speculate beyond their data, but, as witharchitecture, it is quite common to imagine that the changesin women’s dresses would correspond to events outside thefashion system, to wars, economic events or the weather,expressing a certain spirit of their time. What that classicapproach neglects is the degree to which those trajectoriesand their momentums are constrained by the dynamics of thefashion business itself — its techniques of production,marketing and sales — and ultimately by the collectivechanges of taste. And even conceived of as one of Forrester’sdynamic situations, such an account cannot predict when newpossibilities emerge, when women begin to wear pants, forexample, or when some women wear thin skirts, while otherswear wide ones (figure 4.5).

To carry the analogy to its conclusion, new clothingpossibilities emerge at different kinds of thresholds, when thepace of fashion accelerates beyond a certain point or whentoo many women (and men) are participating in the fashionsystem. In his 1960 book on the planning of shoppingcenters, Victor Gruen sought to illustrate the synergisticconditions that enable a new shopping center to emerge, tounderstand the necessary “chain reaction betweeninvestment, income and financing.”20 While the analogy wasdrawn from physics, the dynamics implied are thoroughlyecological. The emergence of a successful shopping center isexplained as a delicate interaction between factors like the“financing climate, economic climate, business potential,management skill, and general cost level.” He used theanalogy and his decades of experience to describe targetvalues for those factors but, of course, this model would onlydescribe the emergence of the form with which he wasfamiliar, not of something different, like big-box retail(figure 4.6).

BIOTECHNIQUES: BUILDING PRODUCT INFORMATIONFor most buildings the critical flows are neither energy norresources, but money and product information. That situationis exemplified by the ever expanding Sweets Catalog and thewhole messy system of selling building materials, productsand processes. Sweets originated in the 1890s as a service ofF.W. Dodge Construction.22 The first full catalog appeared in1906, with an introduction by Thomas Nolan in which he

4.5Dynamic variationsin women’s dressdimensions from1787–1936, byKroeber andRichardson, “ThreeCenturies ofWomen’s DressFashions.”19

4.6Primary andsecondary factors inthe “chain reaction”of shopping malldevelopment, fromGruen and Smith,Shopping TownsUSA, 1960.21

Page 68: B. Kolarevic-Performative Architecture

61

“very gladly consented to commend the idea [of] areally scientific standard catalogue and index ofbuilding materials and construction.” He explainedthat he himself had been working for fifteen years at“finding some practical solution to the ‘CatalogueProblem’ which no architect has been able to work outhimself.” His description of offices overrun with boxes,books and piles of information, and of busy architectswith “less and less time” to do “more and more work”still applies today.23 Although the now multi-volumeSweets Catalog has certainly prospered since 1906,becoming an essential tool in virtually every Americanarchitectural office, the “catalogue problem” has in noway been solved. Like traffic on the highway system,the flow of building information has only increased involume and accelerated in speed with each newimprovement in information technology.

In 1929, a young Danish architect named KnudLönberg-Holm sent an article to the ArchitecturalRecord in which he described the “catalogue problem”as a continuing crisis for the architecture profession,arguing that the solution lay in a radical rethinking ofthe distribution of information in architecture:

… the architect has lost his leadership. From aprofessional man with a professional ethics he hasbecome a business man subject to the whims of thebuyer. The progressive architect acutely realizesthat his problem means ultimately the negation ofhis profession. He has no power to meet hisdilemma through his architectural work. As anindividual businessman he cannot afford theresearch work necessary for the proper executionof his ideas; moreover, he is confronted by the gulfwhich separates him from a client unsympathetictoward an experiment at his expense.24

He argued that “collective problems require collectivethinking and collective work,” and he proposed theinvention of an organization that would act as a“clearing house” and “an economically independentresearch institute,” setting standards and organizing

information. After a brief stint as a technical editor atArchitectural Record, he moved in 1932 to found theresearch office of Sweets Catalog Service. In 1939 he wasjoined in that effort by the Czech designer Ladislav Sutnarand together they reshaped the look and logic of thecatalog, developing the bold graphics and characteristic“S” still used today. Of course Sweets is in no way aneconomically independent institution. It is produced as amulti-volume bound collection of short catalog sectionsprovided by product manufacturers, whose fees andadvertising tie-ins with the Architectural Record andDodge Construction Reports directly support Sweets. As aresult, most of Lönberg-Holm and Sutnar’s work had to beexecuted indirectly by persuading and teachingmanufacturers. They sought to standardize and disciplinetheir advertising inserts, shaping them into documentsreadily used by busy architects seeking information. In thelate 1940s, they formalized their efforts in a pamphletprepared for product manufacturers and that work was sopopular that they brought out an expanded, full colorversion called Catalog Design Progress in 1950. In theintroduction they explained that their aim was to produce“dynamic,” “living standards” that could keep up with therapid pace of technological advance:

Thus with today’s industrial development and theconcurrent higher standards of industry,corresponding advances must be made in thestandards of industrial information itself. The need isnot only for more factual information, but for betterpresentation, with the visual clarity and precisiongained through new design techniques. Fundamentally,this means the development of design patterns capableof transmitting a flow of information…25

Their first section charted the “emergence of new flowpatterns” in all aspects of contemporary life —transportation, production, communication — thendevoted the body of the book to the visual and structuralfeatures with which such information flow patterns shouldbe directed in their catalog. They concluded with a brieftheoretical section that offered “flow” as that form of

Page 69: B. Kolarevic-Performative Architecture

62

information that emerges naturally from the functionaldemands of architectural practice. It was a clever formulationthat overcame the form-function opposition that continues toworry modern architects. They explained the emergentcondition of flow analogically, by comparison with a variety ofother entities newly understood according to the cyberneticconcept of system: “The flow pattern of any sequence adoptsits own form, reflecting function, and its variety of forms maybe observed not only in information flow, but in man (thenervous, digestive, and reproductive systems), in industry(production flow), and elsewhere”26 (figures 4.7 and 4.8).

The management of architectural information by SweetsCatalog has continued with the subsequent migration of theircatalog information onto compact discs in the 1980s and ontothe world wide web in the 1990s, but the original ethic hascontinued: “Comprehensive information correctly formattedand focused on your customer’s needs!”29 In other words, theflow of product information is always channeled according toa powerful network of interests: according to brand identitiesand sales relationships, on the one hand, and to the ever-shifting expression of needs, desires and identities, on theother. What Lönberg-Holm’s original description did notexplain was the degree to which they sought to accelerate thatflow of information and increase the pace of industrialization:

For a continuous advance in production standards theremust also be a continuous liquidation of obsoleteproducts, enterprises, and beliefs. This is possible only inan economy where property relations impose norestrictions on the continuous development of newproductive forces ... This expansion of social wealthimplies increasing industrialization.30

In other words, the system of information flow andindustrialized construction has its own momentum fueled byour individual needs, choices and actions. As many critiqueshave argued, merely fitting better products into normativeconstruction only modulates the effects that industrialdevelopment has on the biosphere. To make a difference, it isnecessary to understand both the structure and velocities ofthe flows already in place, and to locate the threshold effectsthat occur in building.

4.7Sweet’sCatalog File,1949.27

4.8“The flow pattern of any sequence adopts its own form,reflecting function, and its variety of forms may be observednot only in information flow, but in man (the nervous, digestive,and reproductive systems), in industry (production flow), andelsewhere.” From Catalog Design Progress, 1950.28

Page 70: B. Kolarevic-Performative Architecture

63

BIOTECHNIQUES: CUBICLESThe acceleration of biotechniques after World War IIbecame evident in research agendas and the rapiddevelopment of digital technology. And even as highlyrationalized, seemingly mechanized offices were beingbuilt across the United States, the German Quickbornmanagement consulting group were quietly inventing anew form of office layout: the Bürolandschaft or officelandscaping.31 Based on a rigorous analysis ofcommunication patterns within an office, charted throughexhaustive interviewing techniques and diagrams, theydissolved the walls of the office-as-production-line. Theanalogy to a natural landscape was evident in theirpathway diagrams, and in the compelling idea that theform of the office layout was not designed, but emergedfrom the process of analysis. Their detailed diagrams ofcommunication paths and intensities were the tools thatgenerated the landscape plans, which resembled nothingso much as the meandering “desire paths” that animals,savages and undergraduates chart with their feet (figures4.9 and 4.10).

4.9Quickborner officecommunication chart andplan diagram, from FlexibleVerwaltungsbauten.32

4.10Quickbornerorganizationalinteraction diagram,from FlexibleVerwaltungsbauten.33

Page 71: B. Kolarevic-Performative Architecture

64

Those ideas were rapidly communicated throughoutthe planning community and by 1964 the HermanMiller furniture company had formalized them in arevolutionary line of office equipment: Action Office I.Under the guidance of their research director, RobertProbst, they developed the first moveable panels, worksurfaces and storage units that came to define thecubicle and made office landscaping possible. By thelate 1960s, the effects were visible everywhere and theconcept of organic office planning offered a new kindof proportion or regulating system for office layouts:

The rigid patterns of office layout that hadbecome standard during World War I, assumedthe character of time worn tradition by 1960 ...But it failed for precisely that reason. Classicalsystems are inherently inflexible. Since theyembody intellectual-aesthetic ideals of harmonyand order, to disrupt any one element is to destroythe whole. Change is inadmissible. When aclassical order is imposed upon an organic system— one whose parts are related by functions andprocesses that are themselves in flux — the resultis apparent order and actual chaos. An office issuch an organic system. Its organicism, however,is not revealed in those hierarchical charts thatbear so curious a relation to feudal concepts ofthe social orders on earth and in heaven. But,since the actual relations between office personneldefy the caste system codified in charts andembodied in layouts, attitudinal and physicalbarriers were created that seriously blocked linesof communication.34

In close sympathy with structuralist ideas in anthropologyand sociology, and exhibitions like Architecture withoutArchitects and Learning from Las Vegas, the naturalisticforms of Bürolandschaft planning offered anti-authorialdesign strategies that appealed to the generation of 1968.35

As Francis Duffy reported about his own efforts to spreadsuch ideas, “Anthropology with its rigorous comparativetechniques, its search for cross-cultural patterns betweenartifacts, behaviour, societal norms and their technologieswas an obvious model for architectural research. Theinterrelated three-part model of buildings, people andtechnology ... was firmly implanted.”36 Even though theorganic look of the office landscape passed relativelyquickly, the principle of planning around communication,the importance of adaptation and, of course, the cubicle,formed the core of the new biotechniques of the office(figure 4.11).

THRESHOLD EFFECTS: HIGHLYCONDITIONED BUILDINGSIn 1957 the head of the Carrier air conditioning corporationobserved that “whenever 20 percent of the office buildingsin any one city include air conditioning, the remainingbuildings must air-condition to maintain their first classstatus.”38 That process had apparently taken about tenyears, and after the late 1950s it was largely assumed thata high-quality office building in an American city would beconditioned to some degree. The technology had beenavailable for many decades, but it took the particular armsrace dynamic of post-war real estate development to changeit from a desire to a “need.” A similar process had occurredamong movie houses in the 1930s, which along with luxuryhotels had rapidly adopted air conditioning in the pre-warperiod once its competitive advantage had been

4.11Application of officelandscape techniques toDupont administrativeoffices, 1967.37

Page 72: B. Kolarevic-Performative Architecture

65

demonstrated.39 Those examples served to introducethe public to the experience of conditioned air,preparing them for the ever-increasing amounts ofconditioning (figure 4.12).

This is one kind of threshold effect that occurs infeedback systems, when an arms race develops betweencompetitors. They rapidly adopt new products,strategies and quite expensive technologies if theircustomers are free to make other choices. Who wouldgo to a hot movie theater or rent a hot office if a coolone is readily available? And in the process, a new,higher standard emerges and is fixed not only in publicdesires but in normative construction practices andregulatory codes and standards. At that point, the newstandard no longer represents a choice, but a culturallyand officially recognized need. It is not easily reversedand can apparently only be altered by a similarlydynamic cultural process. The energy supply crises ofthe 1970s, for example, temporarily alteredthermostat settings and some social habits, but thelogic of energy conservation quickly receded whenprices dropped.

I do not mean to argue that air conditioning isinherently bad, far from it. The relief from sweatingsimply feels good, and that is precisely why it becomessuch an effective element in competitive situations,leading to a steady escalation of expectations. Theproblems are twofold. The first are very familiar:

greater levels of conditioning produce a whole host ofsecondary environmental effects through heat islandconditions, the use of greater amounts of energy, the releaseof CFCs, and so on. Many of these are amenable to betterdesign or greater efficiency, and form the basis of mostgreen design strategies, but the second kind of problems aremore troublesome. Not only does the escalating aspect ofthis process establish ever higher standards, requiring evergreater levels of conditioning, but the techniques ofconditioning profoundly alter the size and character of thebuildings that can succeed in the marketplace.

In other words, once the real-estate process described in1957 takes place, and conditioning becomes the norm forcommercial buildings, then the scale and configuration ofthose buildings quickly expands so that they have to beconditioned. The dimensions of a commercial buildingdesigned without air conditioning are effectively defined byits external skin, meaning that every inhabited workplacehas to have ready access to a window for light and air. As aresult, even the biggest of the early skyscrapers were madethin by cutbacks, light courts and reentrants. Once theconnection to windows is severed by air conditioning andefficient lighting, the buildings are free to grow (out and up)until they encounter other scale limits: circulation, the sizeof elevators and so on.41 And like the escalation of comfortstandards, this is simultaneously a technical process ofconditioning buildings and a cultural one of conditioning theindividuals who inhabit them (figure 4.13).

4.12A bulky buildingamong other bulkybuildings. The classicfully-conditionedbuilding of the late1950s: SeagramBuilding, 1958.40

4.13Classic, big-but-thin,unconditioned skyscraper.Sullivan, WainwrightBuilding, 1891.42

Page 73: B. Kolarevic-Performative Architecture

66

A building’s balance-point temperature provides arough index of when it crosses that threshold, when itsspaces are no longer directly connected to the outdoorclimate. When a building becomes both sufficientlybig and contains a sufficient intensity of internalconditioning and support systems, its balance-pointtemperature will fall below the average outdoortemperature and it will have to provide cooling forsome part of most days of the year (and everyday intheir windowless cores). This initiates a fairly simplecascade of effects: first air conditioning and efficientfluorescent lighting make it possible to fill largerinterior areas with people and the equipment they useto work, but the people, lights and equipment allproduce heat, which requires even more conditioning.As heat removal becomes ever more important,windows are sealed and are designed to exclude asmuch sunlight as possible, making the interiorenvironment more efficient, but less and less pleasant.

Those two thresholds — higher comfort standardsand bigger buildings — were passed for manybuildings by 1960, establishing the now familiar normfor commercial and retail construction of highly-conditioned buildings with vast interior spaces. But,of course, that norm has been subject to manycriticisms and it has been modified, sometimesradically, in recent decades. Beginning almostimmediately in the early 1960s, there were parallelefforts to introduce green plants and natural light intothe cores of the newly bulky buildings. The plants

initially arrived as part of the office landscape movement(Bürolandschaft) and rapidly found a place in thereinvented (and conditioned) atriums of the late 1960s:the Ford Foundation and the Hyatt Regency of 1968 aretypically cited as the first fully developed examples. Inaddition to its pleasant qualities, the atrium wassubsequently identified as an energy conservationtechnique in the late 1970s and 1980s, and become ahallmark of the higher-quality, more efficient officebuildings of that period (figure 4.14).

The purpose of this thumbnail history of conditionedbuildings is to illustrate the degree to which theenvironmental thresholds important to green design alsoinvolve social and cultural factors, and to explore whythey are so resistant to change. A second kind ofthreshold, one of intensities, is even more critical anddifficult to examine because it involves the whollysubjective experience of the bodies being conditioned.

THRESHOLD EFFECTS: BIOTECHNICAL BODIESThe Environmental Protection Agency (EPA)distinguishes “building related illness,” which can beattributed to an identifiable cause, from sick buildingsyndrome (SBS) in which “occupants experience acutehealth and comfort effects that appear to be linked totime spent in a building, but no specific illness or causecan be identified.”44 The inability to diagnose SBScontinues, though recent epidemiological studies confirmthe correlation between mechanical ventilation rates andreports of SBS symptoms, such as “upper respiratory

4.14Classic, fully-conditioned, atriumbuilding with returnair circulated throughatrium space andplantings. FordFoundation, 1968.43

Page 74: B. Kolarevic-Performative Architecture

67

and mucous membrane symptoms (i.e., irritated eyes,throat, nose, or sinus), and lower respiratory irritation(i.e., difficulty breathing, tight chest, cough, orwheeze).”45 In this regard, SBS belongs to a broadclass of environmental illnesses (EIs), such as multiplechemical sensitivity and Gulf War syndrome, that arewidely reported, but that do not fit any biomedicalexplanation. From one side of the dispute, it is claimedthat such syndromes are wholly somatic, learned groupexpressions of other psychological issues, while on theother side, serious research continues to seek thebiomedical causes and etiologies of the distress.46

What seems evident in both bodies of research isthat the perception of indoor air-quality, of itsfreshness, is central to the syndrome. As the earlyventilation researchers discovered when they firstbegan to investigate ventilation levels in the 1930s,freshness involves both an assessment of the intensityof odors and a judgment about their quality. Like noise,an odor can be pleasant in one situation and offensiveor bothersome in another. What this suggests topsychologically oriented researchers is that sensationssuch as odors can trigger “social psychologicalprocesses of contagion, where complaints andsymptoms spread from person to person, andconvergence, where groups of people develop similarsymptoms at about the same time.”47 From the otherperspective, the remarkable sensitivity of the nosesuggests the possibility of very subtle toxicogenic orallergic processes that have not yet been identified.The statistical correlation between SBS andmechanical ventilation systems, for example, appearsto offer evidence of the underlying physical causesrelated to the rates and processes of ventilation, andhas quickly been acted on by design professionals.48

I can contribute no new evidence or research thatmight resolve the biomedical question, but I wouldargue that as with the previous examples, SBSrepresents the passing of a critical threshold in theconditioning of buildings, a threshold that issimultaneously physical and social. The previousexamples appeared after a certain threshold of scale,

after a certain number of buildings were conditioned orafter a certain size of building was produced, but SBS andother EIs seem to develop at certain thresholds ofintensity. Environmental comfort is defined in these terms,as the intensity of air conditions (temperature, enthalpy,wind, pollution) at which neither our attention nor ourcoping mechanisms are noticeably required. EI sufferersthemselves often explain their symptoms in terms of thecumulative thresholds of toxins or irritants, and they usefeedback system theories to explain the disproportionateeffects that trace amounts of different substances cancause: “total body load, limbic bundling, andhypersensitivity.”49 For designers, it ultimately makeslittle difference whether these are medical or somaticexplanations, they are the point at which systems designedto provide comfort paradoxically begin to threaten thehealth of the occupants with the very intensity of theirconditioning. As a recent sociological study observed, theaccounts of EI sufferers portray “a body that reactsseverely to ordinary commercial furniture designed tooffer it at least a modicum of rest; a body that respondsviolently to air passed through conventional heating andcooling systems designed to make it more comfortable…it is as if this body is in protest against the products ofmodernity and, in its distress, is calling for a radicalchange in the conventional boundaries between safe anddangerous.”50 Environmental illnesses, like SBS, shouldremind us that the real object of environmental design isnot the efficiency of conditioning, but the state of thebodies that occupy them, whose intimate concernscontinue to exceed any performance assessment.

LIVING STANDARDSI have offered this brief outline of biotechniques to maketwo very simple points about the conditioning ofcontemporary buildings. First, environmental conditioningis not just a collection of devices whose performance canbe optimized. They are complex systems that operate onbuildings and people simultaneously, systems with theirown history, trajectory and momentum. Second, there arecritical thresholds in the scale, velocity and intensity ofthat conditioning that radically alter the effects they

Page 75: B. Kolarevic-Performative Architecture

68

produce, meaning that more, or even more efficient,conditioning is not always the answer. In that sense,Kiesler was correct, if technological systems self-organize or evolve according to their own performancecriteria, then the only useful measure of design ishuman health. The concept of health is now largelyassociated with biotechnical medicine, but as SBSillustrates, it still includes social and political forms ofcoping as well. The best term I can offer as a designguideline for healthy thresholds of conditioning is the“living standard” sought by Lönberg-Holm, a standardthat adapts to changing arrangements, and whichallows overly conditioned bodies to actively influencetheir own environments.

To understand what such a living standard mightmean for current practice, architects must look beyondthe narrowly visual terms which have constrained it.Much of the architectural encounter with environmentalconditioning has been devoted to issues of formalexpression. The initial opposition between the

traditional elements of building — walls, windows and roofs— and the wires, pipes, ducts and devices that invaded themin the late nineteenth and early twentieth centuries, gave wayto the “servant” spaces of the Richards Medical Labs, andthen to the vigorous display of service elements at the CentrePompidou. But after fifty years of intensely conditionedbuildings, such debates about the expressive role ofmechanical equipment seem passé.

If we look more closely, however, the history ofarchitectural experimentation reveals a parallel fascinationwith the symbiotic resolution of buildings and machines. FromLe Corbusier’s mur neutralisant (neutral wall) and FrankLloyd Wright’s radiant floors have sprung an entire ecology ofintegrated building components, from the “hairy” and“blistered” skin of Roche and Levaux’s [Un]plug building(figure 4.15) to the ventilating, double-glass façades ofFoster’s Commerzbank. Through such biotechnical elements,buildings are not limited to the symbolic expression of culturalideas, to merely organic forms, but to active demonstrationsof the organic themes that lurk within every mechanism.

4.15An “absorbent”concept building,“hairy” with solarcollector tubes and“blistered” withphotovoltaic cells.Roche and Levaux’s[Un]plug Building,2002.51

Page 76: B. Kolarevic-Performative Architecture

69

NOTES1 “Performance Design” in Progressive Architecture 48,August, 1967, pp. 105–153.2 Frederick Kiesler, “On Correalism and Biotechnique: ADefinition and Test of a New Approach to Building Design” inThe Architectural Record, September, 1939, pp. 60–75.3 Janine M. Benyus, Biomimicry: Innovation Inspired byNature, New York: Harper Collins, 2002; Karel Honzik,“Biotechnics: Functional Design in the Vegetable World” inArchitectural Review 81, January, 1937, pp. 21–22; KarelHonzik, “A Note on Biotechnics” in Circle: InternationalSurvey of Constructive Art, London: Faber and Faber, 1937.4 Kiesler, op.cit., p. 68.5 Ibid.6 Ibid.7 The mechanical and organic analogies have often been flipsides of the same coin. For an insightful discussion see thechapter on mechanical and organic in Luis Fernandez-Galliano,Fire and Memory: On Architecture and Energy, Cambridge:MIT Press, 2000.8 Dennis L. Meadows et al., The Limits of Growth : A Reportfor the Club of Rome’s project on the predicament of mankind,New York: Universe Books, 1972.9 Jay Wright Forrester, Industrial Dynamics, Cambridge, MA:MIT Press, 1961.10 Jay Wright Forrester, Urban Dynamics, Cambridge, MA:MIT Press, 1960.11 Jay Wright Forrester, World Dynamics, Cambridge, MA:Wright-Allen Press, 1971.12 Forrester, op.cit., Industrial Dynamics, p. 191. Fig. 16-4.13 Forrester, op.cit., Industrial Dynamics, pp. 200–201. Fig.16-7.14 Forrester, Ibid., World Dynamics.15 Greg Lynn, “Blogs (or Why Tectonics is Square andTopology is Groovy).” In ANY 14, 1996, pp. 58–61. “AnAdvanced Form of Movement” in After Geometry,Architectural Design 127, 1997, pp. 54–61.16 Greg Lynn, Animate Form, New York: PrincetonArchitectural Press, 1999.17 M. Speaks, “It’s Out There … the Formal Limits of theAmerican Avant-garde” in Hypersurface Architecture,Architectural Design 133, 1998, pp. 26–31.18 A. L. Kroeber and J. Richardson, “Three Centuries ofWomen’s Dress Fashions, a Quantitative Analysis” inAnthropological Records 5:2, 1940, pp. 111–153.19 Ibid.

20 Victor Gruen and Larry Smith, Shopping Towns USA: ThePlanning of Shopping Centers, New York: Reinhold PublishingCorp., 1960.21 Ibid.22 S. Lichtenstein, Editing Architecture: Architectural Recordand the Growth of Modern Architecture 1928-1938, PhDDissertation, Cornell University, 1990.23 T. Nolan, “Introduction” and “Sweets” in Indexed Catalogueof Building Construction, New York: The Architectural RecordCo., 1996.24 K. Lönberg-Holm, “Architecture in the Industrial Age” in Artsand Architecture 84, April, 1967, p. 22.25 K. Lönberg-Holm and L. Sutnar, Catalog Design Progress,New York: Sweets Catalog Service, 1950; W. Sanders, “CatalogDesign,” review of Catalog Design Progress in ArchitecturalRecord, March, 1951, pp. 32 and 35.26 Ibid.27 Ibid.28 Ibid.29 Sweets Catalog Services, 2003: sweets.construction.com30 K. Lönberg-Holm and C. T. Larson, “The Technician on theCultural Front” in Architectural Record, December, 1936, pp.472–473.31 Ottomar Gottschalk, Flexible Verwaltungsbauten: PanungFunktion Flachen Ausbau Einrichtung Kosten Beispiele,Quickborn: Verlag Schnelle, 1968.32 Ibid.33 Ibid.34 “From Grid to Growth” in Progressive Architecture,November, 1969, p. 100.35 Bernard Rudofsky, Architecture without Architects: A ShortIntroduction to Non-Pedigreed Architecture, New York: TheMuseum of Modern Art, 1965; Robert Ventrui, Denise ScottBrown and Steven Izenour, Learning from Las Vegas, Cambridge:The MIT Press, 1972.36 Francis Duffy, The Changing Workplace, London: PhaidonPress, 1992, p. 3.37 Gottschalk, op.cit., p. 221.38 C. Wampler, “Growth of the Air Conditioning Industry” in AirConditioning, Heating, and Ventilating 54, May, 1957; GailCooper, Air Conditioning America: Engineers and the ControlledEnvironment, 1900-1960, Baltimore, MD: Johns Hopkins Press,1998; Marsha E.Ackermann, Cool Comfort: America’s Romancewith Air Conditioning, Washington, DC: Smithsonian InstitutionPress, 2002.

Page 77: B. Kolarevic-Performative Architecture

70

39 Gail Cooper, Air Conditioning America: Engineers and theControlled Environment, 1900-1960, Baltimore, MD: JohnsHopkins Press, 1998; Marsha E. Ackermann, Cool Comfort:America’s Romance with Air Conditioning, Washington, DC:Smithsonian Institution Press, 2002.40 Architectural Review, December, 1958, p. 374.41 This development has been reviewed by a number ofauthors: Solar Energy Research Institute (SERI), The Designof Energy Responsive Commercial Buildings, New York:Wiley-Interscience, 1985; Dean Hawkes, The EnvironmentalTradition: Studies in the Architecture of Environment,London: E & FN Spon; New York: Chapman & Hall, 1996;A. F. C. Sherratt, Air Conditioning: Impact on the BuiltEnvironment, London: Hutchinson; New York: Nichols, 1987.42 Claude L. Robbins, Daylighting: Design and Analysis, NewYork: Van Nostrand Reinhold Company, 1986, p. 5.43 Architectural Record, February, 1968, p. 106.Progressive Architecture, February, 1968, p. 94.44 Environmental Protection Agency, “Indoor Air Facts No.4 (revised): Sick Building Syndrome (SBS)” in Office of Airand Radiation, Office of Research and Development, Office ofRadiation and Indoor Air 9909J, April, 1999.

45 C. A. Erdmann, K. C. Steiner and M. G. Apte, “IndoorCarbon Dioxide Concentrations and Sick Building SyndromeSymptoms” in The Base Study Revisited: Analyses of the 100Building Dataset, Indoor Environment Department, LawrenceBerkeley National Laboratory LBNL-49584, 2002.46 H. Staudenmayer, Environmental Illness: Myth and Reality,Boca Raton, FL: Lewis Publishers, 1999.47 Alan Hedge, “Addressing the Psychological Aspects of IndoorAir Quality” in Asian Indoor Air Quality Seminar, Urumqi,China, September 22–23, 1996.48 O. Seppanen and W. J. Fisk, “Relationship of SBS-Symptomsand Ventilation System Type in Office Buildings” in IndoorEnvironment Department, Lawrence Berkeley NationalLaboratory LBNL-50046, 2002.49 Staudenmayer, op.cit.50 J. Stephen Kroll-Smith and H. Floyd, Bodies in Protest:Environmental Illness and the Struggle over Medical Knowledge,New York: New York University Press, 1997.51 Marc Emery, Innovations Durables: Une autre architecturefrancaise = Appropriate Sustainabilities: New ways in FrenchArchitecture, Paris: Ante Prima; Basel: Birkhauser, 2002.

Page 78: B. Kolarevic-Performative Architecture

71

5

THOMAS HERZOG

PERFORMANCEFORM

Page 79: B. Kolarevic-Performative Architecture

72

5

THOMAS HERZOG

PERFORMANCEFORM

Page 80: B. Kolarevic-Performative Architecture

73

Since its founding more than thirty years ago, thearchitectural practice “Herzog + Partner” has beencommitted to exercising social responsibility throughits projects, while at the same time actively pursuingscientific and technological advances relevant for theenvironment in multiple ways, such as the potential ofharnessing solar energy.

New concepts are developed in collaboration withuniversities and research institutions, such as theFraunhofer Association (Fraunhofergesellschaft), theTechnical University Munich (Technische UniversitätMünchen — TUM) and the Center for Applied EnergyResearch of Bavaria (Zentrum für angewandteEnergieforschung Bayern e.V.). Urban planning pilotprojects, pioneering buildings, and prototypes ofbuilding systems and components are created as aresult of the knowledge gained from these interactions.All of the projects incorporate a special demand foraesthetic quality. The form is not predetermined butcreated, depending on the task, as a result of thedesign process that we call “performance form.”

That working method (the development of“performance form”) is the modus operandi of ourarchitectural practice. The problem and specificmarginal conditions are examined and interpretedsystematically. Different alternatives and solutions areformulated on the basis of the philosophy of thearchitectural practice. In the case of building projects,the local climatic data are recorded in addition to theusual information gathered from investigations, suchas the access and positioning of the building in theurban context.

An important factor, which dominates the scope ofthe practice, is the development of an overallcomposition that includes building structures as well asthe surrounding landscape and public spaces in orderto achieve optimal harmony in the architectural design.Using physical models and computer simulations, theeffects on the form of the buildings, the positioning,and the possibility of using solar energy for heatingpurposes, cooling, ventilation, power generation andcomfort are investigated. The solutions are found

gradually in workshops, together with other members of thedesign team. The client is also involved in this process,making the decision-making process transparent.

SOLAR DESIGNIn the past, the use of solar energy was seen primarily as away to reduce conventional heating energy in buildings andto produce hot water. Great advances have been made inboth areas through constructional developments. The resultscan be realized, among other things, by creating large areasof south-facing glazing and closed, heavily insulated northwalls, or by zoning in the layout of rooms through theorientation of buildings, thereby ensuring a favorablerelationship between the volume and the surface area. Inaddition, technological advances in the fields of heating andhot-water systems now allow a minimum of 60% of hotwater needs in housing to be supplied from solar energy viathermal or storage collectors.

In the 1980s, a disagreement about the use of largeareas of south-facing glazing existed. Solar gains were nottaken into account when calculating the energy balance of abuilding; an improved U-value through adequate thermalinsulation was seen as the best way to reduce fossil fuelconsumption. This came to be recognized as a monocausalview, as it largely ignored the fact that buildings are highlycomplex organisms, functionally, technically andaesthetically.

Even in moderate climates, increased insulation canlead to cooling problems in the summer, especially inadministration buildings. On average, less than 10% of theoverall energy consumption of office buildings isattributable to heating. The energy needs for cooling, on theother hand, represents between 10 and 20%; and coolingrequires roughly three times as much primary energy perKW hour as heating. Today, components with variable g-values (the total energy-transmittance value through amaterial) have become a useful tool in the construction ofexternal walls, allowing buildings to respond differently tochanges in the weather. New solar cooling systems are alsoa promising area of development. Maximum energy poweris available for this purpose when it is most needed. If onereduces the areas of glazing, however, less daylight will

Page 81: B. Kolarevic-Performative Architecture

74

enter a building, thus the proportion of artificial lightingthat is required will be increased. Research has shownthat in Europe the average proportion of energy requiredfor artificial lighting in administration buildings is in therange of 30%.1

Computer tools have long been available forsimulations and data processing in various fields ofconstruction, such as the thermal or lighting balance ofbuildings. Many new components are being developed orare already being tested in these areas. They includevacuum thermal insulation, new types of glass that canreact to the changing needs of insulation, such asswitchable glass with an inert-gas filling and with U-values of around 1.0, electrochromic and thermotropicglass, ventilation components with a solar preheatingfacility, and many other innovations.

Environmental forms of energy can be used for avariety of purposes — for natural lighting, ventilation,heating and the generation of electricity via photovoltaicsystems. As the conditions change according to theseason, the time of day, the weather and the type andduration of use, conflicting needs can arise. One mightexpect a solution to these problems could be provided bythe so-called “intelligent building” systems. Thesesystems respond to changing conditions and can be usedto control different functions, such as heat generation,distribution and output, the operation of sunscreenblinds, co-ordination of daylighting and complementaryartificial lighting, ventilation flaps, and humidifiers.Automatic controls of this kind are largely poweredelectrically and are a common feature in our lives today.Certain reservations, however, can be attributed toextensive automation. These include the vulnerability ofthe technical systems themselves, higher constructioncosts, and a lack of awareness of cause and effect on thepart of users. It is important to understand thephenomena one activates. Electronic systems in buildingsmust, therefore, serve the sense of orientation of thepeople who use them. Human beings should experiencetheir environment — including the artificially createdelements — with all of their senses, in order not to sufferfrom a process of mental and spiritual atrophy.

URBAN CONSIDERATIONSIn the first half of the 1960s, it became clear that thesegregation of urban functions into housing, production andleisure zones led to a loss of quality in modern cities. Atthat time the interest was concentrated on the recovery ofcomplexity. “Concentration, Interconnections, Urbanity”was the title of a seminar held in 1964 at the Departmentfor Urban Planning of the TUM. Younger architects pinnedtheir hopes to large-scale variable structures of the kindthat Yona Friedmann had developed for Paris, EckhardSchulze-Fielitz for the Ruhr area, Kisho Kurokawa forTokyo, and Kenzo Tange for the Tokyo Bay project. In theseschemes, proposals were made for handling massiveincreases in traffic, especially automobile traffic. Thevolume of traffic itself was not in question. Today we knowthat transport accounts for roughly a quarter of all fossilenergy consumption, in addition to being responsible for anumber of negative side-effects. It is necessary, therefore,not only to replace fossil fuels and reduce the volume ofprivate traffic, but to reconsider the causes. Linking urbanfunctions can be a solution because the spatial separation ofindustrial uses from housing is not as necessary today as itwas in the previous century. In concrete terms, this meansthat a mixing of functions should be reintroduced. Urbanstructures that are more neutral in their internal circulationand in their use of space are needed. An additionalimprovement would be to increase building density whereverpossible.

Only in neighborhoods with sufficient purchasing power,and where the facilities serving daily needs are accessibleon foot, is an effective mixture of functions viable; onlyunder these circumstances can vehicular traffic be reducedsignificantly. This would also lead to a reduction of land useand infrastructure costs. The ideal of living in a greensuburbia on the edge of the city may convey a feeling ofproximity to nature, and thus possess a spurious ecologicalappeal. However, it causes dependency on transport and anincrease in fuel consumption and environmental pollution,as well as disastrous social consequences. The lack ofresidential space with urban quality in our city centers isthe cause of the enormous numbers of commuters. Buildingsalso provide too little scope for change. Millions of square

Page 82: B. Kolarevic-Performative Architecture

75

meters of developed space remain unused, preventing access toenergy and material resources. In addition, we allow ourselvesthe luxury of an average of 40 m2 of living area per person inGermany, in comparison to 15 m2 in Italy and much less inMoscow. Anyone who fears that higher densities in cities wouldlead to the insalubrious conditions of the past overlooks theimproved material state of our society.2

GUEST BUILDING FOR THE YOUTHEDUCATION CENTER3

The Premonstratensian monks of Windberg Monastery run ayouth education center for which they commissioned a newdormitory block. The nature and duration of use of the variousgroups of rooms played an important role in determining theenergy concept for the building. The rooms that are used forlonger periods of the day are separated from those that areused briefly. The two sections of the building were alsoconstructed with different materials (figures 5.1 and 5.2).

5.1Site plan, GuestBuilding for the YouthEducation Center,Windberg, Niederbayen,Germany (1987–91),architect ThomasHerzog with PeterBonfig and WalterGötz.

5.2Guest Building,Windberg: upper,ground and lowerground floor plans.

Page 83: B. Kolarevic-Performative Architecture

76

The southern tract houses the lounge areas and thebedrooms, which can be divided in different ways. Therooms on this face are more attractive, with an open view ofthe landscape through broad areas of glazing (figures 5.3and 5.4). A direct exploitation of solar energy and daylightwas also possible for heating and lighting the spaces on thisside, which are used for longer periods. To reducetemperature extremes and to permit storage of thermalenergy, this tract was constructed with heavy, thermallysluggish materials. Opaque areas of the south-facingexternal walls were also clad with translucent thermalinsulation (figure 5.5). This allows solar radiation to passthrough it, but minimizes thermal losses. The south-facingexternal wall is thus heated up during the day and passes onthe thermal energy to the internal spaces after an intervalof five to six hours, beginning in the early evening (figure5.6). In other words, during the night, when externaltemperatures are at their lowest, the outer wall functions asan inward-facing solar heating area (figure 5.8). During thesummer months, overheating is prevented by the broad roofprojection and external louvered blinds.

5.3Guest Building,Windberg: southernfaçade.

5.4Guest Building,Windberg: easternfaçade.

5.5Guest Building,Windberg: cross-section.

Page 84: B. Kolarevic-Performative Architecture

77

The northern tract contains the sanitary facilities andstorage spaces, as well as the circulation route throughthe building. These spaces are distinguished by the factthat they have a generally lower average temperaturelevel, since they are used for only short periods. In theshower rooms, for example, a higher temperature levelis required for only two to three hours a day. This tractwas, therefore, equipped with a quickly functioningwarm-air heating system. To minimize heat lossesthrough ventilation, a heat-recovery unit was installedin the attic space. The hot-water supply is providedlargely from solar energy by means of vacuum-tubecollectors on the roof.

Part of the teaching program of the youth educationcenter is to make the functioning of the buildingcomprehensible to young guests. This is accomplished byproviding them with insights into the use ofenvironmentally sustainable forms of energy through“passive” and “active” constructional systems andthrough the mechanical installations that play a role inthe energy balance. The architectural effect of the newlydeveloped south-facing heating wall is immediatelyvisible in the façade and is tangible internally. Theservice runs, solar storage units and collectors areexposed to view, and a display panel which is installed inthe entrance area shows changes in temperature levels.

DESIGN CENTER4

The congress and exhibition hall in Linz marks a newinterpretation of the concept of early historicalexamples of large glazed halls. These include the CrystalPalace in London (1851–1936) and the Glaspalast inMunich (1854–1931), which provided effectiveprotection against the elements and a hitherto unknowninternal light quality. From the outset, one of the goalsin planning the Design Center in Linz, Austria (1989–93), was to reduce the inner volume of air to aminimum. The internal height of the spaces was limitedto 12 m, and since this clear height was not requiredeverywhere in the hall, the roof structure was designedin a flat arched form with a glazed covering (figures 5.8and 5.9).

5.6Guest Building,Windberg:temperature curvesin south-facingexternal wall on aclear January day.

5.7Guest Building,Windberg: northernfaçade.

Page 85: B. Kolarevic-Performative Architecture

78

The steel girders forming the load-bearing roof structure spana distance of 76 m and cover an area of 16,800 m2. To ensuremaximum flexibility of use, all exhibition and congress spaces(with a capacity of 650 and 1,200 people) adjoin a commonfoyer (figures 5.10 and 5.11). The points of access are laidout in such a way that visitors to concurrent events do notinteract. Continuous longitudinal access routes along bothsides allow the various halls and the gallery space to becombined. The ancillary zones are also laid out in linear form.Since the partitions in these zones can be moved, the spacesremain flexible for the changing uses (figures 5.12 and 5.13).

In developing a natural lighting concept for the building,the challenge was to achieve excellent light quality inexhibition areas without having to make sacrifices in theindoor climate and without giving rise to excessive energyconsumption. In collaboration with the BartenbachLichtLabor, a new kind of building element was developed forthe light-transmitting roof (figure 5.14). A plastic gridintegrated in roof panels with a complex performance allowsindirect luminous radiation from the northern hemisphere ofthe sky to enter the building, while direct sunlight is screenedoff (figure 5.15). In this way, excessive heat gains are avoidedin internal spaces in the summer. Just 16 mm deep, the retro-reflecting grid, thinly coated with pure aluminum, wasinserted into the cavity between the panes of double-glazingover the roof (figure 5.15).

5.8Sprawling DesignCenter, Linz, Austria(1989–93), architectThomas Herzog withHanns Jörg Schradeand Heinz Stögmüller.

5.9Design Center: model.

5.10Design Center:exhibition hall withthe ventilationsystem.

Page 86: B. Kolarevic-Performative Architecture

79

5.11Design Center:congress hall.

5.12Design Center: steelgirders framing theroof structure.

5.13Design Center: endwall with the clay-tile façade.

5.14Design Center:simulation of thelight-transmittingroof.

5.15Design Center:entrance hall.

Page 87: B. Kolarevic-Performative Architecture

80

The geometry for cutting the grid was determined bycomputer programs and had to take account of the followingfactors: the angle of elevation and the azimuth angle of thesun at various seasons, the exposure and orientation of thebuilding, and the slope of the roof. Thermally separated steelsections help to reduce heat losses through the buildingenvelope.

In addition to thermal and daylighting aspects, a specialchallenge was posed by the need to guarantee an adequateair change in this flat, deep building. Fresh air enters viafloor inlets and ventilation flaps at the sides of the hall. Thewarmed, used internal air rises to the top of the building as aresult of thermal buoyancy. During the heating period, theair is then borne by large ducts to a heat recovery plant.During the rest of the year, the exhaust air escapes from thebuilding at the crest of the roof via a large, continuousopening that is fitted with closable louver flaps. To guaranteethe extraction of the vitiated air under unfavorable air-pressure conditions, a “spoiler” capping was developed andassembled over the crown of the roof (figure 5.16). This 7 mwide element has a convex underside and exploits the“Venturi effect” to support the extraction of air from thebuilding (figure 5.17). The final form of this element wasdetermined in wind-tunnel tests (figures 5.18 and 5.19). Inlight of these developments, one sees that building envelopesare subject to changes in their technical functioning andconstruction when, in addition to performing theirtraditional protective role, they are required to controlindoor temperatures and the ingress of daylight.

5.16Design Center: longface of the hall with“Venturi” capping toassist naturalventilation.

5.17Design Center:simulations oftemperature curvesand airflow patterns.

5.18Design Center:simulations usingthe wind tunnel.

5.19Design Center: simulationsusing the wind tunnel.

Page 88: B. Kolarevic-Performative Architecture

81

DEUTSCHE MESSE AG (DMAG)ADMINISTRATION BUILDING5

As a building type, high-rise buildings are not usuallyregarded as compatible with the conservation ofresources. This project proved it was possible to design a“sustainable” building that refutes this opinion. This isaccomplished through a new interpretation of spatial andfunctional concepts, by co-ordinating the form ofconstruction with the energy concept, applying soundprinciples of building physics, and exploiting locallyavailable forms of environmental energy.

Taking account of environmentally relevant issues,high quality in the workplace and flexibility of use werethe main criteria in ensuring that the building could adaptto changing working needs over time (figures 5.20–22).

The layout is articulated into a central working areaof 24 x 24 m in plan and two tower-shaped access cores,containing ancillary spaces, which are offset to the sides.This allows great flexibility in the use of the building,which is twenty stories high. Above the three-storyentrance hall are fourteen floors that are used exclusivelyfor offices. At the top of the building are conference anddiscussion spaces, as well as a story occupied by thecompany management. The individual floors can bedivided into open-plan, combination or single-unit officesas required, whereby a similar quality is guaranteed forevery workplace (figures 5.23–28).

All users can enjoy natural ventilation by opening thesliding casement doors to the intermediate space betweenthe two skins of the façade. When the casements areclosed, fresh air is supplied via inlets from the ventilationducts incorporated in the inner façade skin. Vitiated air isextracted from the offices by means of thermal uplift ofthe warmed air in the internal spaces and is channeledthrough a central duct system, with vertical shafts leading

5.20Aerial view, DeutscheMesse AG (DMAG)AdministrationBuilding, Hanover,Germany (1997–99),architect ThomasHerzog and HannsJörg Schrade withRoland Schneider.

5.21DMAG:conceptualsketch.

5.22DMAG.

Page 89: B. Kolarevic-Performative Architecture

82

up to a rotary heat-exchange unit. In winter, this allows85% of the thermal energy contained in the extractedair to be used for preheating the fresh-air intake.

The integration of thermal storage mass into theoverall concept was of great importance in ensuring anefficient use of energy and a high degree of internalcomfort. The heating and cooling system laid in themonolithic screeds allows the thermal environment to becontrolled at a low temperature level. By storing heatingor cooling energy in the thermo-active floor slabs, whichis released at a later time, it is also possible to reducetemperature extremes, thus ensuring a balanced indoorclimate and agreeable surface temperatures on thespace-enclosing elements.

A distinguishing feature of this structure is the co-ordination of the various building subsystems within anoverall concept. In this way, it was possible to guaranteea high level of comfort with low energy consumption,and to harness sun and wind energy to control the indoorthermal environment and ventilation. A ventilation

tower rises by about 30 m above the northern access core.The exploitation of thermal uplift is an important aspect ofthe natural air-supply and extract system for the entirebuilding (figures 5.29 and 5.30).

The load-bearing structure consists of a reinforcedconcrete skeleton frame with in-situ concrete floors. Thebuilding is braced by the two access towers, which, inconjunction with the floor slabs, form a stable structuralsystem. The access towers are clad with the Moeding façadesystem, a rear-ventilated clay-tile form of constructionsuspended from the main structure.

The double-skin glazed façade to the office areas(figure 5.31) offers several advantages. The glazed outerskin acts as a screen against high-speed winds, therebyallowing natural ventilation (figure 5.32). Sunshading canbe installed in a simple form behind the outer façade layer,where it is protected from the elements and is easilyaccessible for maintenance and cleaning (figure 5.33). Thebuffer effect created by the corridor space between the twofaçade skins (figure 5.34), and the high resistance to

5.23DMAG: groundfloor.

5.24DMAG: standardfloor plan.

5.25DMAG: HermesLounge.

5.26DMAG: boardmembers floorplan.

Page 90: B. Kolarevic-Performative Architecture

83

5.27DMAG:glazing.

5.28DMAG:terrace.

5.29DMAG: the southernand eastern façades.

5.30DMAG: the northernand western façades.

5.31DMAG:double-skinfaçade withfixed glazing.

5.32DMAG: cross-sectionthrough the double-skinfaçade with externalfixed glazing.

5.33DMAG: cross-sectionthrough the double-skinfaçade with externalventilation flaps.

Page 91: B. Kolarevic-Performative Architecture

84

thermal transmission provided by the two layers ofglazing, help to reduce the effects of insulation near thesurface of the inner façade and increase the sense ofcomfort in the rooms. The cantilevered reinforcedconcrete floors and the fire protection they provideallow a form of façade construction with story-highglazed elements. This, in turn, facilitates a maximumexploitation of daylight and creates an ample sense ofinternal space. The cantilevered section of the floor slabdoes not have to be thermally separated from the mainarea as a result of the use of insulating double-glazingin the outer façade skin. In addition, it was possible tolocate the load-bearing columns in the façadeintermediate space where they do not obstruct thefunctional floor area.

CONCLUSIONSArchitecture is one of the few professions with a trulycomprehensive character. Architects have to take aholistic approach to complex systems. It seems likely thatthe methods and patterns of work we know today will besubject to fundamental changes in the coming decades.This applies, in particular, to the nature of co-operationbetween architects and specialists in other fields. A largepart of the problem we face today in respect to thenatural resources is attributable to one-sided processes ofoptimization, in which insufficient attention is paid topotential side-effects. Success comes more easily byignoring disturbing secondary issues than by seeking abalance between the various factors involved. If we are toachieve this balance between technical processes, theharmony of nature and a sense of social responsibility, wemust ensure that a long-term, mutual responsibility forthe public welfare plays a greater role in life.

A society that wishes to develop along humane linesmust be more than the sum of its individuals pursuingtheir own interests. A holistic approach to problems canbe achieved only if we succeed in intensifyinginterdisciplinary thinking in collaboration among the arts,natural and social sciences, engineering and economics,and if environmental design is understood as a complexcentral discipline.

NOTES1 According to investigations carried out by Professor Nick Baker,presented in a lecture in Cambridge, England.2 Adapted from a revised version of a lecture given in the GermanArchitectural Museum in Frankfurt on Main on the occasion of theexhibition “The Ecological Challenge” in February 1997.3 Windberg, Niederbayen, Germany (1987–91), architect ThomasHerzog with Peter Bonfig and Walter Götz.4 Linz, Austria (1989–93), architect Thomas Herzog with HannsJörg Schrade and Heinz Stögmüller.5 Hanover, Germany (1997–99), architects Herzog + Partner,Thomas Herzog and Hanns Jörg Schrade with Roland Schneider.

5.34DMAG: intermediatespace between façadeskins.

Page 92: B. Kolarevic-Performative Architecture

85

6

ALI M. MALKAWI

TOOLS

PERFORMANCE

RESEARCH ANDSIMULATION:

Page 93: B. Kolarevic-Performative Architecture

86

6

ALI M. MALKAWI

TOOLS

PERFORMANCE

RESEARCH ANDSIMULATION:

Page 94: B. Kolarevic-Performative Architecture

87

Simulating building performance requires specializedexpertise that targets the design, engineering,construction, operation and management of buildings.It draws its resources from many diverse disciplines,including physics, mathematics, material science andhuman behavior. Its intention is to predict the behaviorof a building from conception to demolition.

Most of the fundamental work on buildingperformance simulation algorithms and predictions wasdeveloped a few decades ago. Simulation tools havebecome widely available and this has had a measurableinfluence on the way in which buildings are designed,analyzed and constructed. However, building simulationcontinues to evolve and the outcomes of research in thisarea are being incorporated into the design and theconstruction of most buildings, and they are evident inthe recent advances in building simulationenvironments. This progress was driven primarily byfunded research efforts (academic and governmental)and commercial kernel encapsulations, with bothsectors benefiting from advances in computation, suchas new programming paradigms and the increasedpower of computing and the Internet. Funded researchinvolved the development of simulation throughout thelife cycle of the building. Efforts in commercialdevelopment have been primarily purpose driven, takingadvantage of the maturing algorithms which are able toresolve particular issues of building performance. Theseenvironments focus on providing user-friendly interfaceswhile allowing for flexibility in modeling and accuracy.

Advances in building simulation environments havebeen focused on two areas: the structural framework ofthese environments and the activities they support. Thestructure of the simulation involves algorithmdevelopments, data management and interfacing.Simulation algorithms have historically been designedto predict answers to domain-based questions, such aslighting, thermal or structural problems. Each of thesedomains has sub-problems that must be modeled andsimulated differently. These maturing algorithms areunder rapid development in the areas of code validation,uncertainties and efficiency of representations.

To shift the conventional use of such tools from analysisonly to analysis and synthesis, a renewed research intoutilizing advances in optimization is underway. Thisresearch stems from the idea that digital simulation toolscan be used to support performance-driven design usingoptimization and partial automation. Current researchpromises to take advantage of advances in visualizationand human computer interaction developments. It isresponding to the current needs of the design team, whichare far from being fulfilled. Methods to assist in couplingand data management between algorithms to increasetheir prediction accuracy are also being explored. Inaddition, frameworks and standards are being developedto facilitate their integration with other environments inorder to support the design analysis team’s activities. Thefollowing sections in this chapter provide a summary ofthese developments and advances, and discuss thechallenges that exist.1

SINGLE DOMAIN TOOLSMany tools have been developed to predict theperformance of the building in areas such as thermalflows, lighting, acoustics, structures, etc. Becausedifferent problems require different simulationalgorithms, a variety of computational simulation toolsexist. The United States Department of Energy (DOE)maintains an online directory of energy-related tools forbuildings.2 These tools vary from simple, approximateperformance tools to the very precise. The websitedemonstrates the diversity of the tools for solvingproblems of energy-related issues in buildings.

Although refinements and updates have been made tomany of the single domain environmental simulation tools,the two main developments within the past decade are theintroduction of EnergyPlus™ and a surge in the use ofComputational Fluid Dynamics (CFD).

In 1995, the DOE introduced EnergyPlus™ to replaceDOE-2 (sponsored by the DOE) and BLAST (sponsored bythe US Department of Defense), which is based on thebest capabilities of both tools. It launched its first releasein 2001 with a promise of continued support anddevelopment. EnergyPlus™ is a simulation engine without

Page 95: B. Kolarevic-Performative Architecture

88

a user-friendly interface that simulates heat and massenergy flows throughout a building. The concept is toprovide a rigorous objected-oriented environment thatwill attract third-party developers to create userinterfaces and modules, and that can be linked to otherprograms through data and object interaction, asdescribed later in this chapter.3 This concept marks anew era in the way simulation engines will be developedand supported. It also marked a shift from approximate-based simulation and simplified interface developmentfor the non-expert to the support of rigorous enginesthat can be utilized in frameworks and environments toaid in the building design for both the non-expert as wellas the expert user.

As computing power became less expensive, the useof CFD (after evolving for several decades) began toincrease. CFD applies numerical techniques to solve theNavier-Stokes equations for fluid fields and it providesan approach to solve the conservation equations formass, momentum and thermal energy. CFD has beenused in many applications in relation to buildings. Theseinclude natural ventilation design,4 building materialemissions for indoor air-quality assessment,5 andcomplex flows of fire and smoke in buildings.6 Firesimulation using CFD is exemplified by the NationalInstitute of Standards and Technology’s open source FireDynamics Simulator (FDS) (figure 6.1).7 FDS predictsthe smoke and hot air flow of a fire by using Large Eddy

6.1SmokeView —burn room projectshowing 3D smokevisualization.11

Simulation. Using different parameters, CFD is also usedin noise prediction in relation to ducting in buildings.8

Other applications are more complicated and mayintegrate alternative building simulation models, such asEnergyPlus™ described earlier. Although it has beenwidely used, the technology of CFD is still underdevelopment as applied to the solution methods used.9

Studies illustrate the importance of following validationprocedures to ensure the accuracy of the CFD results.10

It is important to highlight that CFD remains an experttool and its use requires knowledge of fluid mechanics toset up the simulation model, populate its boundaryconditions and interpret the results (figures 6.2–6.4).

COLLABORATIVE ENVIRONMENTSAND INTEROPERABILITYIn order to overcome the drawback of needing differentsimulation engines to predict different performanceaspects, integration between algorithms provides ameans to overcome some of these limitations andincreases the efficiency and prediction accuracy ofperformance tools. Research in this area haspredominantly involved issues of inter-model couplingbased on engine and equation integration.12 On the otherhand, as the environments shifted from procedural toobject oriented, simulation-based environments followedthis shift. Object-oriented code that supports modularityand inheritance allowed simulation to be more flexibleand expandable. This shift made it “technically”possible to begin work on encapsulating shared anddistributed simulations. It also made possible thedevelopment of environments and frameworks that canachieve design analysis integration based on semanticrepresentations which support object and datainteraction. This includes custom-driven objectintegration developments, the data and objectinteroperability development, as well as the research inprocess-driven integration.

Custom-driven data integration attempts tointegrate multi-simulation engines into one system thatprovides interoperability between the various tools byusing shared custom data objects. Examples of such

Page 96: B. Kolarevic-Performative Architecture

89

development are the Building Design Advisor (BDA)13

and SEMPER environments.14 Although SEMPERdiffers from the BDA in the way it represents thedesign-analysis activities and its internal interfaces,both use “custom” building model representationsbased on object-oriented technology.15

As interoperability among different softwarebecame a necessity, a collective effort by industry,governmental and research organizations to establishdata exchange standards for the building industryattributed to a surge in object- and data-drivenintegration. This approach targets data sharing amongdifferent software applications, including simulations,to achieve software interoperability. Data sharing isachieved by mapping the relevant data within eachprogram to a generic common data model that containsinformation required by all other programs. Thedevelopment of this common data model is currentlythe mission of the International Alliance ofInteroperability (IAI), which was formed in 1994 andbuilt on the foundation of previous research projects indata exchange for integrated building models andstandards in related areas.16 The view is that thisuniversal framework and specifications for the commondata model — called the Industry Foundation Classes(IFC) — for the architectural, engineering, constructionand facilities management industries can facilitateinteroperability between the existing and future

6.2CFD simulation andanalysis for thePennDesign buildingduring the schematicdesign phase.

6.3CFD simulation andanalysis of CivicHouse, University ofPennsylvania, forrenovation andretrofittingpurposes.

6.4Study of double-skinfaçade (New YorkPolice Station) usingCFD simulation.

Page 97: B. Kolarevic-Performative Architecture

90

software tools used by industry participants17,18

(figure 6.5). Although the latest version of the IFC(IFC2x 2nd Edition) captures significant datainformation of the complete building model, the goalof promoters and developers to have specificationsfor every class of building element, required tosupport data-sharing between all software tools usedby building professionals, poses a challenge as it isan extensive task to build such a comprehensive datamodel.

To achieve interoperability among simulationtools during the building design and constructionstages, the process-driven integration approachsuggests there must be more than just dataexchanges supported by a common buildingrepresentation. Performance simulation needs aframework which allows it to be called upon at theright time and for the right design decision. Theavailability of domain-specific tools by themselvesrequires additional functionalities and frameworks in

order to play a better role in design evolution. Thisapproach focuses on the effective use of performancesimulation tools in the design process, with fullparticipation from the design team including the expertconsultants. In 2001, a team of researchers fromGeorgia Institute of Technology, Carnegie MellonUniversity, and the University of Pennsylvania beganwork on the Design Analysis Integration Initiative (DAI)project in an attempt to develop credible solutions forthe integration of building performance analysis tools inthe building design process.20 It intended to capitalize onthe efforts already invested in the development ofbuilding product models — the IAI–IFC effort describedabove — without making limiting assumptions about thedesign process or the logic of the design analysisinteraction flow. This process-driven integrationapproach is still in its early stages and will require moreresearch and development to illustrate its full potentialto solve the problem of integration.

PERFORMANCE TECHNIQUES AND METHODSTo increase the utility of the simulation in design, severalmethods and techniques are being incorporated. Decisionsupport environments and optimization in the models, aswell as user-friendly interfaces combined withvisualization techniques, are the main areas ofdevelopment in regard to performance-based simulationfor architectural design.

Research in optimization and decision supportenvironments began two decades ago. Computationalalgorithms were used to develop systems that assisteddesigners in their activities by providing either guidancethrough advice or optimization using emerging ArtificialIntelligence techniques. Knowledge-based systems andcomplex problem-solving methods were researched fortheir potential use as an aid for both expert and non-experts to perform and interpret simulations.21-24 Theircontribution has been significant in initiating the use ofcomputational techniques to solve performance-baseddecision-making problems. In addition, this research hasinspired renewed efforts in the areas of optimization and

6.5Industry FoundationClasses — life-cycleworkflow structure.19

Page 98: B. Kolarevic-Performative Architecture

91

decision support environments. This work is dominatedby the idea that design is a goal-oriented decision-making model, where goals are defined by desiredperformance values. The work benefited from rapidimprovements in the field of optimization to solvecomplex problems in areas of numerical methods,solution strategies and the development of newalgorithms.

Design optimization using performance simulationhas been researched recently to test the applicability ofspecific algorithms in regard to their effectiveness.During the past few years, stochastic methods such assimulated annealing and genetic algorithms (GAs) havebecome prevalent. They have been applied to a widerange of problems, including thermal and lightingperformance in buildings.25,26

Recently, investigations of gradient-based andderivative-free methods were also conducted. Gradient-based methods are efficient and their results are reliableexcept in cases of complex simulation.27,28 On the otherhand, derivative-free deterministic methods perform wellfor problems that suffer from simulation noise and aremostly suitable for small problems.29 Because simulation-based optimization can be time consuming, approximate-based methods which use functions derived fromsimulation responses to partially guide the search duringthe optimization process have been recently utilized.30-33

Strategies that combine two or more methods havealso been used to overcome problems associated with oneparticular method. An example of such an approach iscombining genetic algorithms and pattern searches toderive a hybrid method to reduce computational run timein problems involving expensive simulations.34

Although optimization algorithms have been used ina variety of ways in building simulation, their fullpotential in design synthesis has not been fulfilled. Thework conducted over the past few years demonstratesthat design optimization can be used to improve buildingperformance and provides rigor in the way the simulationtools can be utilized. An example of such an approach ispresented in the following section.

DESIGN ANALYSIS VERSUSPERFORMANCE-BASED SIMULATIONFundamental to simulation tool development are questionsrelated to the user and the problem the tools need to solve.The answers will orient the development to take a certaindirection. Currently, simulation tools are analysis-basedand support a variety of design activities. Rethinking theuse of these tools from analysis to performance-basedactive design support has not been explored fully. Inaddition, the full potential of the recent developments inproblem-solving techniques and visualization has not beenrealized. Incorporating such issues has the potential toallow active design support to exist. Taking advantage ofthese developments, as well as partially automating someof the simulation processes, will afford simulation tools alarger role in design activities.

An illustrative example of such research is a projectwe recently conducted that integrated computational fluiddynamics and GAs.35 The project uses thermal andventilation performance criteria as a means to generatecreative design alternatives. This process requires an in-depth knowledge of the problem and considerableexpertise. Each design change made to space geometry orits boundaries requires the user to re-model, re-mesh andsubsequently re-compute the airflow. The trade-offbetween different design changes generally remainsobscured because of the complexity of the model. As aresult, the design space is difficult to exploresystematically and solutions can be overlooked.Evolutionary techniques coupled with visualization wereused as a solution to such a problem.

Evolutionary algorithms have traditionally been usedto solve optimization problems. In addition, they can beused as a design aid. The evolutionary approach is a“generate-and-test” approach that corresponds well to theprocedures for design synthesis and evaluation in thedesign process. In an evolutionary-based generativeprocess, design representations are specified as a set ofparameters and as a corresponding set of constraints.Generative design describes a broad class of design wherethe design instances are created automatically from a

Page 99: B. Kolarevic-Performative Architecture

92

high-level specification. Design evolutions can be usedas an aid in stimulating the designer creativity. Theadvantage of such an evolutionary approach is thecreation of diverse sections of the state space that meetperformance targets and increase the possibility fordiscovering a variety of potential solutions by providinga larger search space for designers to interact.

The project demonstrated the potential ofexploring and visualizing the design evolution and itsform generation based on a set of performance targets.It demonstrated how a performance-based designevolution model allows efficient exploration of designalternatives using a four-layer approach: designevolution, performance evaluation, morph visualizationand design evaluation. The design evolution uses a GAmodule that generates the shape of the designinstances. To explore the design space using GAs, thedesign variables are represented as an array of allelesets or “space genome”. For each design change, aCFD analysis is automatically performed to evaluatethe thermal and ventilation efficiency performance,and the results are fed back into the GA module. Thisprocess runs recursively until the best designs arereturned. To capture the system’s solutions and in-between instances as the solution is evolving, thesystem is integrated with a morphing module. This

module allows the diverse instances of designs to bevisualized along with their performance in order toenhance the user’s potential for design discovery and toprovide aid within the design process. During morphing,users can intervene to stop the process and select aninstance based on its form (figure 6.6). This designinstance can then be evaluated for its performance usingan integrated CFD engine. By integrating an optimizationmodule that creates discrete instances of design with amorphing module, the system was able to provide anexample of continuous evolution of optimization.Although reversing the cycle from the design-analysisparadigm to a performance-based simulation-decision-support paradigm has been previously researched, the fullpotential of this shift has not been realized. Despite theadvances in computational techniques that can be used inthe development of such tools, this area is still in itsinfancy. This is partly due to the multidisciplinaryexpertise required for such development and thefragmentation of the building industry, which makes itdifficult to establish a unified voice to articulate suchrequests.

DATA VISUALIZATION ANDINTERFACE DESIGNMost of the available tools provide a one- or two-dimensional representation of the data derived from abuilding performance simulation. This has always been animportant challenge, as only experts can preciselyunderstand the data and hence are always required tointerpret them. Consequently, this introduces theproblems of time and cost, not only in terms of hiringthese experts but also in establishing communicationamong the participants. This communication is not onlydependent on their physical presence, it also involvesissues of representation as well as of semantics.Advances in visualization led to new developments insimulation. Different technologies have made it possibleto create environments that are virtual or augmented.Although immersive building simulation is still in itsresearch and development stage, virtual and augmentedenvironments have been used in a variety of areas in

6.6Performance-based DesignEvolutionsoftwareinterface.

Page 100: B. Kolarevic-Performative Architecture

93

relation to buildings. This includes the extension ofvisual perception by enabling the user to see throughor into objects,36 such as maintenance support forvisualizing electrical wires in a wall or constructiongrids.37 Other applications include structural systemvisualization,38 augmented outdoor visualization39 anda collaborative design process.40

In the area of immersive building simulation,41

only a few projects have been developed — some ofwhich are related to the post-processing of CFDdata,42,43 augmented simulations,44 building and datarepresentation,45 building performancevisualization,46,47 immersive visualization forstructural analysis,48 and interactive immersivebuilding simulation49,50 (figure 6.7).

The few studies conducted in this area illustratethe potential application of such research. Despite theprevalence of challenges related to software, hardwareand the knowledge required to apply this to thebuilding design, immersive environments provideopportunities which are not available using currentsimulation models and interactions, and it extends onthe three-dimensional performance informationvisualization of buildings. It presents a new way ofinterfacing with the built environment and controllingits behavior in real time. This will become increasingly

evident as additional techniques (e.g. optimization) lendmore power to these environments as users navigatethrough them and interact with their elements.

CONCLUSIONSThe use of performance simulation in architecturaldesign is on the rise. This is due mainly to the increaseof computing power and the maturing of the buildingsimulation field. The full integration of simulationwithin the design process is far from complete. However,the building industry, including architects, is aware ofthe need for the better integration of these tools into thelife cycle of the building. Many of the developmentspresented in this chapter are a testimony to the rapiduse of these tools, methods, and technology into design.However, only elite practices are currently takingadvantage of the recent developments in performancesimulation.

The tools are predominately of analysis and are notfor analysis and synthesis. Challenges related toperformance simulation accuracy, which is influenced byfactors such as user interpretations and interventions tovariations in simulation variables and behavioraluncertainties and validations, are still beinginvestigated. Although a collective effort to developstandards for integration is underway, the chapterillustrates that this is a non-trivial issue. As the demandto advance performance simulation increases, newparadigms have begun to emerge and will influencebuilding design and construction. These paradigms areindications of the field’s various directions, some ofwhich are purely demand-oriented, while others forecastthe needs of the field in an attempt to enhance it. Thischapter has illustrated examples of such paradigmswhich include issues of process-driven interoperability,the influence of optimization techniques into the designof new tools, and new means of interaction between theusers and simulations. Although such work is underway,we are far from utilizing the full potential of theadvancement in the computational fields that canimpact the way architecture is practiced andexperienced.

6.7Immersivebuildingsimulation.

Page 101: B. Kolarevic-Performative Architecture

94

NOTES1 A. M. Malkawi, “Developments in EnvironmentalPerformance Simulation” in Automation in ConstructionJournal 13:4, 2004, pp. 437–445.2 DOE online directory of energy related tools for buildings canbe accessed at: http://www.eren.doe.gov/buildings/tools_directory/3 D. B. Crawley, L. K. Lawrie, C. O. Pedersen, F. C.Winkelmann, M. J. Witte, R. K. Strand, R. J. Liesen, W. F.Buhl, Huang, R. H. Henninger, J. Glazer, D. E. Fisher and D.Shirey, “EnergyPlus: New, Capable and Linked” in Journal ofArchitectural and Planning Research 21:4, 2004.4 G. C. da Graca, Q. Chen, L. R. Glicksman and L. K. Norford,“Simulation and Wind-driven Ventilative Cooling Systems foran Apartment Building in Beijing and Shanghai” in Energy andBuildings 34: 1, 2002, pp. 1–11.5 H. Huang and F. Haghighat. “Modeling of Volatile OrganicCompounds Emission from Dry Building Materials” in Buildingand Environment 37: 12, 2002, pp. 1349–1360.6 S. M. Lo, K. K. Yuen, W. Z. Lu and D. H. Chen. “A CFDStudy of Buoyancy Effects on Smoke Spread in a Refuge Floorof a High-rise Building” in Journal of Fire Sciences 20: 6,2002, pp. 439–463.7 G. Forney, D. Madrzkowski and K. B. McGrattan.“Understanding Fire and Smoke Flow Through Modeling andVisualization” in IEEE Computer Graphics and Applications,Los Alamitos, CA: IEEE Publications, 2003.8 Fluent Inc., Fluent User’s Guide, New Haven, USA, 2003.9 Y. Jiang and Q. Chen. “Effect of Fluctuating Wind Directionon Cross Natural Ventilation in Building from Large EddySimulation” in Building and Environment 37: 4, 2002, pp.379–386.10 Q. Chen and Z. Zhai, “The use of CFD Tools for IndoorEnvironmental Design” in A. M. Malkawi and G. Augenbroe(eds), Advanced Building Simulation, London: Taylor &Francis, 2004.11 Forney et al., op.cit.12 Malkawi, op.cit., “Developments in EnvironmentalPerformance Simulation.”13 K. Papamichael, J. La Porta and H. Chauvet, “BuildingDesign Advisor: Automated Integration of Multiple SimulationTools” in Automation in Construction Journal 6: 4, 1997, pp.341–352.14 N. H. Wong and A. Mahdavi, “Automated Generation ofNodal Representations for Complex Building Geometries in theSEMPER Environment” in Automation in ConstructionJournal 10: 1, 2000, pp. 141–153.

15 Malkawi, op.cit., “Developments in EnvironmentalPerformance Simulation.”16 ISO-STEP, ISO 10303 — the industrial automation systemsand integration standards can be accessed at: http://www.iso.ch17 The IAI can be accessed at: http://www.iai-international.org/iai_international18 V. Bazjanac and D. B. Crawley, “Industry Foundation Classesand Interoperable Commercial Software in Support of Design ofEnergy-efficient Buildings” in Proceedings of BuildingSimulation, IBPSA, 1999, pp. 661–668.19 IAI, op.cit.20 G. Augenbroe, P. de Wilde, H. J. Moon, A. Malkawi, R.Brahme and R. Choudhary, “The Design Analysis Integration(DAI) Initiative” in Proceedings of Building Simulation,Eindhoven, The Netherlands: IBPSA Publications, 2003.21 A. D. Radford and J. S. Gero, Design by Optimization inArchitecture and Building, New York: Van Nostrand Reinhold,1988.22 J. Pohl, L. Myers and A. Chapman, “A Knowledge-basedSystem for the Design of Passive Solar Buildings” in ASHRAETransactions 96: 2, 1990.23 E. Shaviv and Y. E. Kalay, “Combined Procedural andHeuristic Method to Energy-conscious Building Design andEvaluation” in Y. E. Kalay, Evaluating and Predicting DesignPerformance, New York: Wiley, 1992.24 A.M. Malkawi, Building Energy Design and Optimization:Intelligent Computer-aided Thermal Design, PhD dissertation,Georgia Institute of Technology, 1994.25 J. A. Wright, H. A. Loosemore and R. Farmani, “Optimizationof Building Thermal Design and Control by Multi-criterion GeneticAlgorithm” in Energy and Buildings 34: 9, 2002, pp. 959–972.26 R, Choudhary and A. M. Malkawi, “Integration of CFD andGenetic Algorithms” in Proceedings of the Eighth InternationalConference on Air Distribution in Rooms, Technical University ofDenmark and Danvak, Denmark, 2002.27 H. Jedrzejuk and W. Marks, “Optimization of Shape andFunctional Structure of Buildings as well as Heat SourceUtilization Example” in Building and Environment 37, 2002, pp.1249–1253.28 R. Choudhary, A. M. Malkawi and P. Y. Palambros, “AHierarchical Design Optimization Framework for BuildingPerformance Analysis” in Proceedings of Building Simulation,Eindhoven, The Netherlands: IBPSA Publications, 2003.29 M. Whetter and E. Polak, “A Convergent Optimization Methodusing Pattern Search Algorithms with Adaptive PrecisionSimulation” in Proceedings of Building Simulation, Eindhoven,The Netherlands: IBPSA Publications, 2003.

Page 102: B. Kolarevic-Performative Architecture

95

30 Choudhary et al., op. cit., “A Hierarchical DesignOptimization Framework for Building Performance Analysis.”31 M. Whetter and E. Polak, op.cit.32 A. M. Malkawi, R. S. Srinivasan, Y. K. Yi, and R. Choudhary,“Performance-based Design Evolution: The Use of GeneticAlgorithms and CFD” in Proceedings of Building Simulation,Eindhoven, The Netherlands: IBPSA Publications, 2003.33 R. Choudhary, “A Hierarchical Optimization Framework forSimulation-based Architectural Design,” PhD dissertation,University of Michigan, 2004.34 M. Whetter and J. Wright, “Comparison of a GeneralizedPattern Search and a Genetic Algorithm Optimization Method”in Proceedings of Building Simulation, IBPSA, 2003.35 Malkawi et al., op.cit., “Performance-based DesignEvolution: The Use of Genetic Algorithms and CFD.”36 G. Klinker, D. Stricker, D. Reiners, R. Koch and L. Van-Gool,“The Use of Reality Models in Augmented Reality Applications.3D Structures for Multiple Images of Large ScaleEnvironments” in Proceedings of the European WorkshopSMILE, 1998, pp. 275–289.37 A. Retik, G. Mair, R. Fryer and D. McGregor, “IntegratingVirtual Reality and Telepresence to Remotely MonitorConstruction Sites, a ViRTUE Project” in Artificial Intelligencein Structural Engineering, 1998, p. 459.38 S. Fiener, A. Webster, T. Kruger, B. MacIntyre and E. Keller,“Architectural Anatomy” in Presence, no. 4, 1995, pp. 318–325.39 M. O. Berger, B. Wrobel-Dautcourt, S. Petitjean and G.Simon, “Mixing Synthetic and Video Images of an OutdoorUrban Environment” in Machine Vision and Applications, no.11, 1999, pp. 145–15940 P. Frost and P. Warren, “Virtual Reality used in aCollaborative Architectural Design Process” in Proceedings ofInformation Visualization, IEEE International Conference,2000, pp. 568–573.41 A. M. Malkawi, “Immersive Building Simulation” in A. M.Malkawi and G. Augenbroe (eds), Advanced BuildingSimulation, London: Taylor and Francis, 2004.42 V. J. Shahnawaz, J. M. Vance, and S. V. Kutti,“Visualization of Post-processed CFD Data in a VirtualEnvironment” in Proceedings of DETC99, 1999, pp. 1–7.43 A. M. Malkawi and E. Primikiri, “Visualizing BuildingPerformance in a Multi-user Virtual Environment” inProceedings of the ACSA International Conference, 2002.44 A. M. Malkawi and R. Choudhary, “Visualizing the SensedEnvironment in the Real World” in Journal of Human-Environment Systems, no. 3, 1999, pp. 61–69.

45 M. Pilgrim, D. Bouchlaghem, D. Loveday and M. Holmes, “AMixed Reality System for Building Form and Data Representation”in IEEE, 2001, pp. 369–375.46 A. M. Malkawi and R. Choudhary, op.cit., “Visualizing theSensed in the Real World.”47 Pilgrim et al., op.cit.48 N. Rangaraju and M. Terk, “Framework for ImmersiveVisualization of Building Analysis Data” in InformationVisualization IEEE International Conference, Los Alamitos, CA:IEEE Publications, 2001, pp. 37–42.49 A. M. Malkawi, R. S. Srinivasan, B. Jackson, Y. K. Yi, K. H.Chan and S. Angelov, “Interactive, Immersive Visualization forIndoor Environments: Use of Augmented Reality, Human-ComputerInteraction and Building Simulation,” Information VisualizationIEEE International Conference, 2004.50 A. M. Malkawi and R. S. Srinivasan, “Multimodal Human-Computer Interaction for Immersive Visualization: IntegratingSpeech-gesture Recognitions and Augmented Reality for IndoorEnvironments,” IASTED Computer Graphics and Imaging (CGIM)Conference, IASTED, Canada, 2004 (in press).

Page 103: B. Kolarevic-Performative Architecture

96

Page 104: B. Kolarevic-Performative Architecture

97

7

GODFRIED AUGENBROE

PERFORMANCEDIALOGUES

A FRAMEWORK

BUILDINGFOR RATIONAL

Page 105: B. Kolarevic-Performative Architecture

98

7

GODFRIED AUGENBROE

PERFORMANCEDIALOGUES

A FRAMEWORK

BUILDINGFOR RATIONAL

Page 106: B. Kolarevic-Performative Architecture

99

This chapter deals with an engineering perspective on thedevelopment and maintenance of buildings. Performanceplays a major role in the expectations expressed by ownersand occupants, and the fulfillment of them by designers andbuilding operators. The “disconnects” between expectationsand fulfillment are rampant throughout the buildingdelivery process, and the better matching of the two isconsidered an important target of the building industry, inorder to become more client-driven and to provide bettervalue overall thereby guaranteeing customer satisfaction.1

The need to move the industry in this direction hasfostered the introduction of a performance-based buildingmethod.2 These methods focus on the generation ofperformance-based statements of requirements and on themanagement of a transparent process that guarantees theirfulfillment. Essentially, this requires better methods andtools to support the communication between designers,engineers and building managers. An essential part of thiscommunication is the “dialogue” that explicitlycommunicates expectations and fulfillments between ademand and supply party. Two dialogues are of particularimportance: (1) the architecture and engineering (A/E)procurement dialogue, which deals with the way servicesare procured by the design team to engineer buildingsystems that meet functional needs and client expectations;and (2) the tenant/facility manager dialogue, which dealswith the proper maintenance and management of thefacility in a way that meets the expectations of theoccupant, owner or portfolio manager and that provides andmaintains maximum value from the facility for allstakeholders.

Little elaboration is needed on the fact that the currentways and methods to maintain and support these dialoguesare hampered by a range of deficiencies which need to beovercome. Among these deficiencies we highlight the“asymmetric ignorance” between the communicatingpartners, the lack of transparent requester-provider roles inthe process,3 the lack of objectively quantifiable expressionsof requirements, and a lack of the proper assessment toolsto ascertain whether expectations have been fulfilled by aproposed design (in dialogue 1) or by a proposed tenant/building allocation (in dialogue 2).

In addition to the failure of the community at large to addquantified elements to the dialogues, there is also anapparent cultural resistance, mainly from the side of thearchitectural designer, where many seem to believe thatarchitectural performance cannot be measured. Thehabitual conjecture in this discussion is that many aspectsof performance can only be interpreted based onqualitative judgments. It is also debated whether thesejudgments have to rely on unpredictable manifestations ofthe design in its not fully determinable context of use.Unfortunately, these judgments are biased by the “valuesystem” of the one who measures them and, at best, lead tosome measure of quality, i.e. “a set of characteristics thatare perceived to contribute to value.”4 In the followingsections it is argued that this is not good enough and thatmany aspects of buildings can and should be measuredobjectively. The performance characteristics that are mostamenable to an objective statement are those that relate tothe functions that the building, or one of its (sub)systems,are designed to perform. Instead of subjective quality,objective “utility” should be introduced to represent anaggregation of objectively measurable performancecharacteristics. The aggregation is performed over systemsand functions.

Traditionally, the dialogues mentioned above have beencast in prescriptive terms, i.e. specifying aspects of thesolution rather than the expected performance of thesolution. Building codes and regulations have longcontributed to this by basing their approach on prescriptivespecification methods. However, this is now no longer thecase, as many countries are moving parts of theirregulations and standards to the performance domain.5

Different methods have come into existence to supportthe dialogue between the demand and supply side. Aninteresting method which is in use today is documented inthe set of ASTM (American Society for Testing andMaterials) standards for Whole Building Functionality andServiceability.6 The standards lay out a methodology knownas “Serviceability Tools and Methods” (ST&M) which canbe used through various stages of the project deliveryprocess. ST&M provides a method to measure how well adesign proposal, or an existing facility, meets the

Page 107: B. Kolarevic-Performative Architecture

100

requirements of the stakeholders, for about 100different performance topics. The current ASTMstandards provide a broad-brush, macro-level method,appropriate for strategic, overall decision-making. Theydeal both with demand (occupant requirements) andsupply (serviceability of buildings).7 The expression ofdemand and supply is captured per topic in so-calledscales, each ranging from the lowest score (0) to thehighest score (9). The scores reflect a rating of the(design of a) facility, both from the demand as well asthe supply perspective. What is rated on the supply sideis the “fitness” of the facility to meet a given need,hence the term serviceability. More information can befound in the work done by Szigeti and Davis.8

Whereas the ST&M method provides a coarsemethod geared towards the corporate use (many of thetopics relate to the serviceability of organizationalneeds), and is based on a scoring by trained users, weare focusing on the introduction of a fine-grained set ofperformance quantifiers that rely solely on metricsderived from first order physical principles, i.e. rulingout the interpretation of experts. We argue that indoing so, an expanding set of performance metrics isadded as quantified elements in the expectation-fulfillment dialogue. One must realize thatquantifications are only useful when they are veryprecisely defined; otherwise they lead tomisinterpretation and do not support a rationaldialogue.

This chapter introduces a framework for therational expressions of building performance, based onthe notion of objectively quantifiable measures. It willdiscuss how these measures can be viewed as a set ofuniquely defined “performance indicators,” whosequantifications are linked to observable buildingbehavior in specific (virtual) experiments. Thequantification perspective in this chapter ispredominantly engineering driven. A system-theoreticalbasis is explored to express both requirements andassessed performances at varying levels of buildingsystems granularity, and its potential to support designdialogues is verified.

The treatment is positioned to support rational decision-making during the different stages of building delivery anduse. The focus is specifically on the fulfillment of clientexpectations during design evolution and on the assessmentof building tenant combinations over the service life of thebuilding. A toolkit for the rapid quantification of measuresfor new and existing buildings is introduced.

The following sections will discuss the methods todefine performance metrics that lead to objective andrational elements in the dialogues addressed above. Thefirst step towards this is a framework that enables theidentification of systems and their contribution to desiredfunctions, and this is the subject of the next section.

A BUILDING SYSTEMSCLASSIFICATION FRAMEWORKThe rationalization of building performance in dialoguesbetween stakeholders requires the systematic discovery ofperformance domains and their specialization to the levelwhere individual measures of performance can beaddressed. Such individual measures of performance willbe termed performance indicators (PIs).

There is no established theory for the discovery anddefinition of PIs. Wim Gielingh in his General AECReference Model9 hypothesizes a design process in whichthe design problem is reduced into smaller manageablefunctional units (FUs). For each of these FUs, a designerlooks for a technical solution (TS) that satisfies thefunctional requirements of the FU. When no solution isfound, the design problem is restated with a new set offunctional requirements, or a new solution is developed tomeet the functional requirements. Solutions to morecomplex design problems are obtained by reducing thedesign problem into smaller design problems. These smallerdesign problems are themselves new FUs that require newTSs. This iterative process continues until solutions havebeen found for all FUs. PIs can be thought of as the“meat” between the FUs and TSs in Gielingh’s model(figure 7.1). Although Gielingh’s work was ground-breaking in its conceptualization, it was too “idealized”and has, as a result, stayed far removed from practicalapplications.

Page 108: B. Kolarevic-Performative Architecture

101

Another systems approach10 to building design dividesthe activity into two distinct but related views: (1) thefunctional view, which addresses the functionalaspects of the building design; and (2) the buildingsystems view, which deals with systems that aredesigned or selected to fulfill the functionalrequirements of the building. The main function of abuilding can be decomposed into lower-level functions,such as safety, habitability and sustainability. Subsetsof these lower-level building sub-functions have beenidentified by various standards bodies, such as theInternational Organization for Standardization (ISO).The lower half of figure 7.1 shows the aggregation oftechnical systems from subsystems, until we reach thelevel of the well know major “functional systems,”such as lighting, fire safety, etc. These systemscontribute to one or more of the required buildingfunctions. Elements are the smallest decomposition ofa building system and may belong to one or moresubsystems. There exists a constraint relationshipwithin elements of a system and between elements ofdifferent subsystems. This systems view is illustratedin figure 7.1.

The goal of the performance approach to building designand maintenance is to provide a mechanism by which onecan measure how well a building system is capable offulfilling the functional requirements of the building thatit serves. In order to do this, the performance approach tobuilding design addresses building requirements in termsof performance rather than properties. In other words,based on test results rather than prescribed dimensions,specific attributes or specific materials. This allows anymaterial or product that meets the performancerequirements to be considered for use.11 There are sevenmain concepts in the performance approach (figure7.2).12,13

1. Goal (occupancy) is usually a qualitative statementthat addresses the needs of the user-consumer anddetermines the required level of performance.

2. Functional requirements are the mandatoryrequirements that must be fulfilled to ensure usersare satisfied with the facility.

3. Performance requirements are the user requirementsexpressed in terms of the performance of a product.Performance requirements are measured by PIs.

4. PIs are quantifiable indicators that adequatelyrepresent a particular performance requirement. API, by definition, is an agreed-upon indicator that canbe quantified using a verification method.

5. Verification methods are used to evaluate whetherthe performance requirement has been met.Verification methods can be experiments (tests),calculations, or a combination of both. Each PI hasits own verification method. A number of verificationmethods may exist for measuring the sameperformance requirement.

6. Functional elements are parts of a building thatfulfill one or more functions to meet the userrequirements. Functional elements can be materials,construction, products from a product catalogue, andeven non-physical entities such as building spaces.

7. Agents alter the ability of functional elements tosatisfy the functional requirements.

7.1Overview of afunctional systemsapproach.

7.2The relationshipamong concepts inthe performanceapproach.14

Performance Aspects

Sub-systems

Elements

Building Function

Sub-building Functions

Aggregated Technicalsystems

Internal Relationships[Constraints]

BUILDING FUNCTIONS

BUILDING SYSTEMS

Func

tion

al U

nits

Tech

nica

l Sol

utio

ns

TS

FU

FU FU FU

TS TS

TS TSTS

GARM Model

FunctionalRequirement

FunctionalElement

Agent (Stress)

PerformanceRequirement

Occupancy(Use)

delegated_to

affects

evaluated_by

dictated_by

affects

PerformanceIndicator

VerificationMethod

tested_by

measured_by

computed_by

Page 109: B. Kolarevic-Performative Architecture

102

The first two concepts provide the functional goalsthat must be achieved. The relationship between theperformance aspects and the building systems definesthe level of performance that is required to achievethat goal. PIs are used to quantify the performancerequirement in order to determine whether aparticular design satisfies the performancerequirement. The context information required by theverification method is provided by the functionalsystem that represents all the properties and all therelationships of all the elements in that system.

Although these existing frameworks provide auseful model for the top-down discovery of PIs, theyare not specific enough to be deployed in the generalbuilding case. One of the main reasons for this is thatno standard decomposition of functional requirementsand no set of ready-made subsystems exist in thebottom half of figure 7.1. With the infinite variabilitythat characterizes the current style of building designmanagement, the framework does not offer a usefulway out. A more bottom-up oriented approach to thediscovery of PIs can provide a better way to achievemore precision in the mentioned dialogue. Designershave opportunities to use the performance approach atall stages of the design process. Not only does theperformance approach enable the designer todiscipline his own contribution but, perhaps moreimportantly, it provides the potential for a clearerdefinition of responsibilities when the analysis part ofthe process is delegated.

Such a bottom-up approach is by nature domain-specific,dealing with a specific analysis expertise and a limited set offunctional requirements and a subset of the systems thatcontribute to that functionality. In the next sections, we willfocus on functional requirements concerning energy usage,thermal comfort and daylight usage. This requires focusingon those building systems that contribute to these functions,such as the heating and cooling systems and the buildingenclosure system. As figure 7.1 demonstrates, these buildingsubsystems can be viewed as compositions of standardcomponents which are not going to be redesigned in everybuilding. It is at the connections of systems and functionsthat PIs are introduced as a way to express their contributionto reaching a desired functionality. It is important to pointout that there is no unique way to measure the contributionof a particular system to an overall function. Many differentways to measure performance (i.e. in different experiments)can be introduced, each with its own merit. So, eachconnection in figure 7.1 may have multiple PIs attachedto it.

A practical bottom-up discovery of PIs may beattempted by closely investigating the type of stakeholderinteractions with respect to expected and fulfilledperformances that take place in real projects. Different PIscan be found in different types of performance evaluationrequests.

Although the above approach will reveal the primaryconnections between functions and systems, and thus pointtowards the need for one or more PIs that “objectify” thisrelationship, we will need a conceptual framework to definethe PIs themselves. This is the subject of the next section.

PERFORMANCE INDICATORSThis section introduces the concept of Analysis Functions(AFs) as a formal quantification method of a PI.

Figure 7.3 shows the basic notion of an AF as a mappingof experimental input variables, environmental and controlvariables and system properties (p) to a quantified PIthrough a specific aggregation procedure and illustrates allthe components that are part of a full AF model. As an AFmodel uniquely describes the input and usage variables, thePI is dependent on the system properties, i.e. PI(p).

The experiment Aggregation

Observable

States PI

Environmental

Conditions and Controls

Input

The system (object of the experiment)

characterized by attribute set p

7.3The components of theAF description.

Page 110: B. Kolarevic-Performative Architecture

103

Many types of performance (e.g. thermal comfort) canbe measured in different ways, and can be usedinterchangeably by the expert community, e.g. (i) thenumber of hours per year that the room air temperatureexceeds a certain threshold, or (ii) the yearly PredictedMean Vote (PMV) distribution.15 Both methods ofmeasuring thermal comfort performance intend tomeasure the same performance aspect, and set up thesame virtual experiment but use different metrics (adifferent aggregation method leads to differentmetrics). This results in the definition of two distinctlydifferent PIs. However, in this case, the two PIs wouldrequire the same virtual experiment (in all likelihoodperformed by the same software) but would usedifferent ways of aggregating the outputs to quantifythe two different measures.

Elaborating on the thermal comfort performanceexample, one can say that thermal comfortperformance is delivered by the “comfort controlsystem,” composed of the heating, cooling, control(thermostats and control system), and enclosuresystems. Different quantifications differ in the way thattemporal temperature information is translated into ameasure, i.e. in the choice of aggregation method. Thedifferent PIs are based on the same type of experiment:a human is placed in a certain location in a given spaceof the building in the local climate. The experimentvariables determine thermostat control, ventilationactions (opening of windows) and observer properties,such as the activity level and clothing, as suggested infigure 7.4.

It should be noted that the AF must be described formallyin order to rule out any judgment or interpretation issuesduring the quantification. This formal description will bereferred to as the “AF model,” constituting a fulldescription of the experiment. The AF model for a thermalcomfort PI formally describes all the entities that areneeded to calculate the relevant state variables thatdetermine thermal comfort. In the case of PMV, there needsto be an internal air zone that has an average airtemperature, and there needs to be surfaces that havetemperatures that can be used to calculate a mean radianttemperature. Also, occupants need to be defined that havea metabolic rate and clothing value. However, if thedecision had been made to base the thermal comfortanalysis function on a different measure, for instance theuse of degree hours for the air temperature, then therewould not have been a need to include any occupant oroccupant properties, and the treatment of the surfacesmight have been different. Note that different AFs forthermal comfort, like a PMV-based and a degree hour-based function, can coexist in the dialogue, and they shouldboth be defined independently of their realization.

Going back to the different types of experiments thatmay be deployed, a couple of observations can be made.First, the mapping from (p) to behavior of the system is thekey part of the quantification of the PI. It should be clearthat the theoretical foundation of this mapping determinesthe reliability of a PI. In the case of a real experiment, theexperimental set-up must be such that all disturbances arekept to a minimum and the monitoring noise does notsignificantly influence the measured behavior (as expressedthrough gathered state information). In a virtualexperiment, one must be able to guarantee that the(simulation) tool’s representation is adequate in order toaccurately predict the behavior of the system. In thoughtexperiments, biases must be avoided by clearly statedprocedures. This can be accomplished by developingprocedures that map the experimental variables onto arating scale. Examples of these normative procedures arethe already introduced serviceability rating methodsdeveloped by Davis and Szigetti16 and the sustainabilityrating method LEED.17 All the experiments provide PI(p)

Reference weather and control actions including user operations, thermostat settings, operation regime of operable

windows, shading etc.

The experiment: Simulate space with one occupant in

reference climate

Aggregation 1 based on T air

Aggregation 2 based on PMV

Person, activities,

clothing etc Observable states

The system : One space with identical neighbors

Parameters : Enclosure elements, material, properties,

furniture air supply etc.

PI 1

PI 2

Reference weather and control actions including user operations, thermostat settings, operation regime of operable

windows, shading etc.

The experiment: Simulate space with one occupant in

reference climate

Aggregation 1 based on T air

Aggregation 2 based on PMV

Person, activities,

clothing etc Observable states

The system : One space with identical neighbors

Parameters : Enclosure elements, material, properties,

furniture air supply etc.

PI 1

PI 2

7.4AF realizations forthermal comfort PI.

,

Page 111: B. Kolarevic-Performative Architecture

104

mapping based on a set of rules that describe theexperiment in precise terms in order to perform thefollowing two steps without any ambiguity:

Step 1: Perform experiment which results inObject_state (experiment_variables, object (p); t).

Step 2: Perform time and space aggregation overObject_State, resulting in PI(experiment_variables, p).

Based on a formal description of the measure (theabove two steps) in an AF model, it will becomepossible to automate the execution of the AF. For this itis necessary to translate the AF model into a formalcomputer readable representation. This representationshould capture the object (geometry, materialproperties, etc.), the experimental conditions andcontrol settings, as well as the aggregation procedures.A data modeling language can adequately capture thisinformation. Recent work that we conducted18 shows anexample of a workbench to accomplish this, usingEXPRESS-G as a graphical language which allowsentities and their relationships to be presented. Thediagrams can be translated to an EXPRESS model, acomputer interpretable version of EXPRESS-G, used inthe implementation stages of the computer systems.The conceptual models structure all the necessary andsufficient information of the experiment.

The selection of which PIs should be embedded in aperformance vocabulary is ultimately a choice ofelegance and efficiency, and should be driven by thesound application of the 80-20 rule, which suggeststhat one aims at a relatively small set of PIs thatrequires moderate effort to develop and which stillcovers most of the (routine) dialogues in A/Eprocurement. We assume that a manageable set ofconcise PIs can be designed with that purpose in mind.

As this section suggests, the derivability of anecessary and sufficient set of PIs is as yet unproven inpractice. Hence there is no guarantee that a limitedand manageable set can be derived to express the

majority of PIs needed in rational decision-making. A morepractical approach is to develop toolkits that quantify adedicated small set of PIs that have particular significanceto a design office, a client, a tenant organization orcorporate owner. The next section shows an example of thistype of development for a tenant organization withemphasis on portfolio management.

APPLICATION: A PERFORMANCE TOOLKITFOR PORTFOLIO MANAGEMENTThe need for continuous monitoring of the technicalperformance of buildings over their lifetime becomesobvious if one realizes that buildings undergo drasticchanges and redefinition of their internal client processes.These changes are caused by internal reorganizations,refurbishments, change of tenants, etc. The naturaldegradation of the technical systems is another reason totrack the performance of the assets in the portfolio and tolink it to maintenance scheduling.

Using the approach of the previous sections, a set ofperformance measures was developed that provide cost-effective, quantitative predictions and assessments of howwell buildings enable specific client functions.

The first version of the toolkit provides metrics forperformance tracking in the areas of energy, lighting,thermal comfort and maintenance. The metrics areembodied in a set of PIs based on standardized andnormative calculation routines that assess a buildingquickly, reflecting its actual use. Based on the theories inbiophysics and physiology, these measures quantify theperformance of a building system in producing a desiredcondition, related to an activity or need of the tenant orany other stakeholder. The metrics allow the evaluator tounderstand how multiple systems interact to produce agiven level of building performance. The resulting PIs canbe used to formulate requirements as well as to quantifyactual performance.

The set of PIs have been harnessed in an operationalbuilding performance assessment toolkit specificallyadapted for the use by tenant organizations for portfoliomanagement. The toolkit is, however, equally relevant todesign firms in the A/E procurement stages and for

Page 112: B. Kolarevic-Performative Architecture

105

corporate owners who realize that buildings should beconstantly monitored to guarantee a safe, healthy andproductive environment for their occupants. For agovernment organization such as the General ServicesAdministration (GSA), the use of the toolkit will deliver thedata that underpin the day-to-day service management aspart of GSA’s tactical operations. Figure 7.5 shows how thebuildings in the portfolio will be assessed on a regular basis(e.g. once a year). The gathered data will be entered and PIquantifications will be done automatically by the toolkitroutines. The data will then be uploaded in a database wherePIs are stored per category. At this point in time, the toolkitcontains roughly 25 indicators. The uploaded data can bebrowsed and linked to other data about the monitored assets(pictures, floor plans, maintenance records, and others). Theaccess to the records of the facility over previous yearsprovides a way to inspect deterioration trends or to inspectsudden changes in performance, e.g. when a new tenant isallocated to the building. As should be clear from the above,PIs measure the building tenant combination, as the virtualexperiment measures behavior relative to the functions andneeds of an occupant. As figure 7.5 implies, all performancedata are accessible for total building quality management.All other sources of data, such as local physicalmeasurements, post-occupancy evaluation (POE) data, as wellas design programs, ST&M records, etc., will be linked toenable data mining for the business intelligence gatheringthat is necessary to improve the design development,procurement and facility maintenance procedures.

Until now, the work has concentrated on fourperformance aspects: energy, lighting, thermal comfort, andmaintenance. Some of the issues involved in generating theindicators and their benchmarking on real buildings areaddressed below.

The list of Pis is shown in table 7.1.

Aspect Function PI Meaning

Energy Energy PI 1–7 Heating, cooling, humidifying, lighting, pumps, fans, hot water in MJ

PI 1 Electric lighting energy consumption over required illuminance level in kWh/m2 • year • lux

PI 2 Luminous efficacy of luminaires in LER (Lumens/watt) Energy efficiency

PI 3 Daylighting autonomy: percentage of hours without requiring an artificial lighting

PI 4 Ratio of task illuminance as installed and as required

PI 5 Outward visibility (view to outside): percentage of occupants who can see the outside from their workplaces

PI 6 Daylighting glare avoidance: percentage of office hours in discomfort range (Daylighting Glare Index • 24, just uncomfortable)

Lighting

Visual comfort

PI 7 Shading devices for glare avoidance (under development)

Air diffusion PI 1 Percentage of occupants in comfort in ADPI (Air Diffusion Performance Index)

PI 2 Hourly average Predicted Percentage of Dissatisfied occupants (PPD) during office hours over a year

PI 3 Percentage of hours where the PPD is in the comfort range (10%)

Asymmetric thermal radiation due to hot/cold glazing

PI 4 Average of PPD where the PPD is not in the comfort range

PI 5 Hourly average PPD during office hours over a year

PI 6 Percentage of hours where the PPD is in the comfort range (10%)

Cold draft caused by glazing

PI 7 Average of PPD where the PPD is not in the comfort range

Diversity PI 8 Average PPD of workers in different activities and clothing levels

PI 9 Heating airflows variation in different rooms in a thermal zone Zoning

PI 10 Cooling airflows variation in different rooms in a thermal zone

Thermal comfort

System’s capacity and response time

PI 11 The seconds required to increase the zone temperature by 1˚C at the peak load time

PI 1 Building Performance Indicator (BPI), scaled from 0 to 100 Efficiency

PI 2 Maintenance Efficiency Indicator (MEI)

PI 3 Manpower Sources Diagram (MSD): a ratio of in-house and outsourcing expenditures

PI 4 Managerial Span of Control (MSC): a ratio of a manager and subordinated personnel

PI 5 Business availability in per cent: an available floor area over an entire floor area over year

PI 6 Manpower Utilization Index (MUI) in per cent: a ratio of man-hours spent on maintenance and total available man-hours

Business and organization

PI 7 Preventive Maintenance Ratio (PMR) in %: a ratio of man-hours spent on preventive maintenance and spent on total maintenance (preventive + corrective)

PI 8 Urgent Repair Request Index (URI): occurrence/10,000 m2 Timeliness

PI 9 Failure frequency: occurrence/10,000 m2 and unit repairing time in minutes

Mainte-nance

Policy PI 10 Maintenance productivity: state/$ (under development)

7.5The landscape ofbuilding performanceassessment in largeorganizations.

Table 7.1Performance indicators (PIs).

Page 113: B. Kolarevic-Performative Architecture

106

To elucidate the development issues of the indicators,we will again take the example of thermal comfort. Forits assessment, eleven Pis were developed to account forspace air diffusion, asymmetric radiation, cold draft,occupant diversity, HVAC zoning and system’s response(refer to table 7.1). It is important to note that all ofthese effects can be studied through full simulation ofthe office spaces. This, however, would defeat thepurpose of a rapid repetitive evaluation. A fullsimulation would also introduce bias into the PIquantification, as no simulation can be performedwithout introducing assumptions and simplifications,often dependent on the type of simulation tool. Theremedy against this bias is the introduction of“normative” calculation procedures. These proceduresare derived such that the indicators are objectivemeasures of a certain performance aspect. Although theresulting value may not always be interpretable as anabsolute (in physical terms) measure for the observedperformance, the approach is ideal for comparativestudies. An even bigger advantage of a normativelydeclared PI is the fact that its value is directly relatedto the relevant set of building and client parameters.

Taking the reasoning from the previous section, anormative indicator is a measure that is definedthrough a simplified but indicative experiment. Thesimplification of the experiment allows the derivation ofand aggregation over the output state of the experimentto be expressed as closed equations. These equationsreside as spreadsheets or formulae in the toolkit,whereas the toolkit user is automatically prompted forall the needed building and client parameters.

In the case of thermal comfort performance, thechoice of necessary PIs was driven by the followingobservations. Thermal discomfort occurs frequently forone of the following reasons: (1) the required averagetemperature is not maintained by the HVAC system; (2)the air supply system creates discomfort zones close todiffusers; (3) discomfort due to asymmetric exposure tocolder façade elements; (4) cold drafts in the perimeterzone; (5) combining different dress codes and activitiesin one zone, creating too much comfort range diversity

in the same zone; (6) inadequate zoning of the HVAC supplyand control system; and (7) the inadequate capacity of theHVAC system under maximum load conditions. In actualsituations the total discomfort will result from thesuperposition of all of these phenomena. Looking at it thisway, it is not surprising that thermal discomfort is stillcomplaint number one in office buildings! The novelty of thetoolkit is its capacity to deal with all these phenomenaseparately, i.e. by introducing one or more PIs for each of thediscomfort “agents,” as indeed table 7.1 suggests. Each ofthe PIs is evaluated based on a simple experiment leading toa closed form calculation, some of which are based onexisting normative calculation procedures. The Air DiffusionPerformance Index (ADPI), for instance, was originallydeveloped as a design guideline for selecting appropriate airsupply devices. In our case it is used as PI1, a measure oftemperature variation, air mixing and the presence ofobjectionable drafts.

To account for the asymmetric thermal radiation, whichis one of most common reasons for discomfort in perimeterzones in offices, three PIs (PI2, PI3, PI4) are introduced inthe toolkit. PI2, PI3 and PI4 are based on the PMV and thePredicted Percentage of Dissatisfied occupants (PPD). Thevalue is normatively calculated from the glazing surfacetemperature and the corresponding Mean RadiantTemperature (MRT) of the façade. The calculation routine isbased on the diurnal and seasonal hourly weather variationsobtained from TMY2 data records (Typical MeteorologicalYears). The hourly glazing temperature is calculated by usinga steady-state calculation method with information on the U-values of 48 different types of glazing systems provided byASHRAE19 and the corresponding mean radiant temperatureis estimated in accordance with ASHRAE Standard 55.20 Theother PIs in the thermal comfort category are all developedalong similar lines.

It should not come as a surprise to the reader that anormatively declared PI calculation has to be tested carefullyin order to ascertain that it is indeed a good and monotonicindicator of the considered performance. Monotonic behavioris especially important as one needs to ascertain that anincrease (in this case the optimal PI value is zero) in the PIvalue leads indeed to a monotonic increase in discomfort.

Page 114: B. Kolarevic-Performative Architecture

107

Of course one of the immediate benefits of a set of theintroduced set of measures is its unequivocal use instatements of client expectations and its use in rapiddesign evaluation. Owing to the direct relationship todesign and occupant parameters, one can directlyinspect the influence of certain design choices.

The toolkit is currently benchmarked in ten officebuildings in the south east of the United States. Onecase of the benchmarking in the area of energyperformance is reported below.

As one of the targets of the energy PIbenchmarking, the Sam Nunn Atlanta Federal Center(AFC) in Atlanta, Georgia, was chosen. The AFC (figure7.6) is one of the largest federal office buildings on theeast coast and provides office facilities for multiplefederal agencies. The AFC consists of four connectedbuildings (a 24-story tower, a 6-story bridge, a 10-story mid-rise tower and a 6-story building). The totalbuilding is served by five chillers. The total floor areais about 148,000 m2 and the total number ofoccupants is approximately 4,500.

For data gathering and survey, three visits weremade to the building. First, information such as floorarea, wall area, window area and U-values wascalculated from available building records. Then,relevant system information, such as the HVAC systemand its operation scheduling, HVAC fan powers (kW),etc., were obtained through the GSA facility manager.

After filling the toolkit data entry form with allrelevant data items, the energy PIs show up as revealedin table 7.2.

In order to make a comparison with real consumption data,the computations were carried out with the actual Atlantaclimate for 2000 and 2001. Note that the normativecalculations are based on the standard reference year forAtlanta, enabling a year to year comparison of the energyperformance, not biased by the annual weather variations.For benchmarking purposes, we have to compare PI valueswith actual use. In the case of the energy PIs, this iswarranted as the normative procedure gives reliableestimates of the expected real energy consumption.

A big advantage of the normative calculation is that itprovides a breakdown of energy consumers, as shown intable 7.2. The outcome of the breakdown will help buildingowners and facility managers to have a sense of where theattention should be paid and where the budget should beallocated to improve energy efficiency.

Owing to the lack of energy sub-metering in the SamNunn building, the breakdown in table 7.1 could not becompared with metered data on-site. But it should be wellunderstood that this (diagnostic use) is not the primepurpose of the PI. Rather, it defines the performance of thedifferent energy consumers, enabling comparison acrossbuildings of the energy consumed by each of the categories.It will enable the building owner to define the minimumperformances for the lighting or cooling contingent oncertain client and building types. This will lead to a muchmore refined “energy service management” based on the PIinstruments than currently offered by consumptionmeasurement (with or without sub-metering).

The differences between normative and actual totalyearly energy consumption in the year of 2000 and 2001are 2.2% and 15.4%. The deviations occur mostly duringthe winter season (figure 7.7). As explained before, thedeviations between the normative expectation and the realenergy consumption are attributable to occurrences thatare not part of the “as designed” operation of the building.

Performance indicator

Energy consumer Average in MJ (2000–01)

PI1 Heating 8,553,494 7.6%

PI2 Fans for ventilation and air circulation

9,090,290 8.0%

PI3 Lighting 54,949,602 48.6%

PI4 Pumps 4,579,134 4.0%

PI5 Cooling 30,225,076 26.7%

PI6 Humidifying 4,006,742 3.5%

PI7 Domestic hot water 1,691,157 1.5%

Total 113,095,495 100.0%

7.6Aerial view of theSam Nunn AtlantaFederal Center (AFC),Atlanta, USA (1997–98), architect KohnPederson FoxAssociates (thepreviously existingRich Annex datesfrom 1924).

Table 7.2Breakdown ofenergy consumersin the Sam Nunnbuilding.

7.7Comparison of thecalculation andenergy consumption(MJ)

Page 115: B. Kolarevic-Performative Architecture

108

In this case the reasons for the deviations becameobvious during an on-site inspection. The two mainreasons were: (1) the full-time (including nighttime)operation of the HVAC system to prevent freezing (thiswas outside the use specification of the system andtherefore not considered in the normative calculation);and (2) the use of personal heaters, mostly by workersin the perimeter zones of the office spaces (thethermostat settings of the “as designed” operationproved to be too low to provide sufficient levels ofthermal comfort).

The AFC did not obtain Energy-Star labeling overthe last two years. Such a label is given to buildingsthat are more energy efficient than 75% of their“peers,” based on actual utility billing. The AFCcurrently ranks only in the 62 percentile on the Energy-Star office scale, and the energy bill in 2001 was about$1.6 million. In order for the building to obtain Energy-Star labeling, an 18% reduction of energy consumptionwould be required. By using the toolkit, a simplesensitivity study was easily made by changing severalparameters, showing how the building’s Energy-Starrating could be improved from 62 to 68, as shown intable 7.3. As exemplified here, the “normative” natureof the toolkit makes it readily usable for a sensitivity/feasibility study of: (1) new buildings in the designstage; or (2) existing buildings to be refurbished.

Note that even with a combination of design(orientation and windows ratio) and operation changes(lighting controls), the building would not obtain theexpected Energy-Star rating of 75. In the currentbuilding only a drastic redesign of the lighting controls(including task lighting) could possibly accomplish this.

The ease with which this recommendation wasgenerated, based on the use of the toolkit, is regardedas a strong endorsement for the development andexpansion of the PI approach. The benchmarking of theother PIs shows similar encouraging results.

CONCLUSIONThe introduction of PIs as unbiased measures ofperformance has great potential for enhancing thedialogue between stakeholders in the building deliveryand management process. It has been shown how asystem-theoretic approach may enable the systematicintroduction of PIs. This way, a growing set of quantifiersof building performance are offered that could be infusedin more rational decision-making between buildingparties, such as in the dialogues between client andarchitect, between architect and engineer, and betweentenant and portfolio manager. The introduction oftoolkits to “operationalize” the PIs is an important stepin this infusion process.

A number of important research questions need to beaddressed before large-scale toolkits will emerge:

1. Can a limited and distinct set of PIs be defined thatcover a significant percentage of recurringdialogues?

2. Can the approach capture performance at increasinglevels of granularity in accordance with designevolution or will the necessary number of PIs“explode”? The answer to this question will bedetermined largely by the establishment of a PIclassification and the way it can be managed fordifferent building systems on different levels ofgranularity.

3. Will rationalization lead to the capture of bestpractice in current building/engineering design andthus be able to act as a catalyst for re-engineering?The building performance analysis profession isgaining in maturity but lacks clear standards andaccepted quality assurance methods. The diffusion ofbest practices could prove to be an important factorfor this maturation and could foster the furtherdevelopment of performance measures.

From To Savings (MJ) $/year

Orientation South East–North West

South–North 1,373,109 16,877

Windows ratio 40% 30% 3,783,711 46,507

Lighting Central on/off Daylight switch 10,989,920 135,082

Total 16,020,313 196,913

Table 7.3Obtainable annualenergy savingscalculation.

Page 116: B. Kolarevic-Performative Architecture

109

NOTES1 Lauri Koskela, An exploration towards a production theoryand its application to construction, Dissertation, Espoo, Finland:VTT Building Technology, VTT Publications 408, 2000, p. 296.2 G. C. Foliente, R. H. Leicester and L. Pham, “Development ofthe CIB Proactive Program on Performance-based BuildingCodes and Standards,” BCE Doc 98/232, Australia:International Council for Research and Innovation in Buildingand Construction (CIB), 1998.3 Jeffrey L. Beard, Edward C. Wundram and Michael C.Loulakis, Design-Build: Planning Through Development, NewYork: McGraw-Hill, 2001.4 Godfried Augenbroe, “Performance aspects of buildingsystems design,” Graduate course syllabus, Georgia Tech,College of Architecture, 2004.5 Greg Foliente, “Developments in Performance-Based BuildingCodes and Standards,” in Forest and Products Journal 50: 7/8,2000.6 ASTM, “ASTM Standards on Whole Building Functionalityand Serviceability,” West Conshohocken, PA: ASTM, 2000.7 Francoise Szigeti and Gerald Davis, “Matching People andTheir Facilities: Using the ASTM/ANSI Standards on WholeBuilding Functionality and Serviceability,” in CIB WorldBuilding Congress, April 2001, Wellington, New Zealand, 2001.8 Ibid.9 W. Gielingh, “General Reference Model for AEC ProductData,” ISO TC184/SC4 Document Number 3.2.2.3. Delft,Netherlands: ISO, 1987.

10 H. A. J. Ridder, Organization of Complex Problem SolvingProcesses, PhD Thesis, TU Delft, Delft, 1996.11 J. G. Gross, “Developments in the Application of thePerformance Concept in Building” in Rachel Becker and MonicaPaciuk (eds), 3rd International Symposium: Applications of thePerformance Concept in Building 1, Tel Aviv, Israel: The NationalBuilding Research Institute, 1996, pp. I-1 to I-11.12 Foliente et al. (1998), op. cit.13 D. J. Vanier, M. A. Lacasse and A. Parsons, “Modeling of UserRequirements Using Product Modeling” in Rachel Becker andMonica Paciuk (eds), 3rd International Symposium: Applications ofthe Performance Concept in Building 2, Tel Aviv, Israel: TheNational Building Research Institute, 1996, pp. 6-73 to 6-82.14 Ibid.15 P. O. Fanger, Thermal Comfort, Copenhagen, Denmark: DanishTechnical Press, 1970.16 ST&M, Serviceability Tools and Methods, ASTM standard inGerald Davis and Francoise Szigetti, ASTM, E. 7, 2001.17 US Green Building Council (USGBC), LEED Rating System,version 2.0, 2001.18 Godfried Augenbroe, Ali Malkawi and Pieter de Wilde, “APerformance-based Language for Design Analysis Dialogues” inJournal of Architectural and Planning Research, 2004 (in press).19 ASHRAE, ASHRAE Fundamentals, Atlanta, GA: ASHRAE,2001.20 ASHRAE, “Thermal environmental conditions for humanoccupancy,” ANSI/ASHRAE Standard 55-1992, Atlanta, GA:ASHRAE, 1992.

Page 117: B. Kolarevic-Performative Architecture

110

Page 118: B. Kolarevic-Performative Architecture

111

8

CRAIG SCHWITTER

BASED DESIGNIN USE

ENGINEERING

PERFORMANCE-COMPLEXITY:

Page 119: B. Kolarevic-Performative Architecture

112

8

BASED DESIGNIN USE

ENGINEERING

PERFORMANCE-COMPLEXITY:

CRAIG SCHWITTER

Page 120: B. Kolarevic-Performative Architecture

113

Pressure for speed, economics and reliability incommercial building and infrastructure design hastraditionally led engineers to rely on standard, code-based prescriptive approaches to solutions for structuraland environmental systems. But as architecturalsolutions become increasingly complex, in part due to thepower of expanding arsenals of architectural softwareand the intense spread of architectural information andconstruction technologies on an international basis,engineers must continue to offer solutions that enable theboundaries of building design to expand and flourish.

New developments in performance-based design offera rapid and technical approach to unique systemsanalysis — methods that in the hands of designengineers, applied at the right time in the design process,can develop quality building design solutions well beyondthe typical and prevalent prescriptive practices in usetoday. Building and infrastructure engineers are seeingrapid changes in the tools we use — changes that canlead to enhanced design capabilities and more creativethinking under the right circumstances. The architecturalcommunity is keenly aware of these developments and isbeginning to capitalize on the emergence of newcomputational tools in building design. Structuralengineering, HVAC engineering, fire and life safetydesign, and construction engineering are all poised tocapitalize on the emergence of new computational toolsin the future — a future for architecture that will havedifferent market demands, design codes and technicalchallenges than we can imagine today.

CHANGE AND INNOVATIONEngineers may often resist it, but design is really allabout change. While the evolution of building practicerelies on the predictability and constant nature of scienceand materials, it also requires us to constantly developnew design approaches for an ever evolving architecturaland infrastructure landscape. But how do we asengineers change and innovate? With a culture that islittered with litigation, and one that is slow to adoptchange, the construction industry is hardly the place tolook for allies in this effort.

However, change is inevitable — a part of fundamentalhuman development. It is a simple fact that the buildingengineering solutions of today will become outdated overtime. As engineers, we can contribute significantly to theeffort for change but often we do not. Engineers are moretypically reliant on “rule of thumb” design or processes thatare familiar and proven — an understandable pattern,especially as an engineer’s view must be conservative attimes. But what about intellectual curiosity and thecreativity of engineering design? What about using “whatif” more often? A stubborn attitude to change is theengineering profession’s heaviest albatross, and yet, at itsmost fundamental practice, engineering design, or theapplication of science, has potentially the most to offer inthe evolution of the built form.

Even as we strive for change, we must look back asmuch as we look forward. Prescriptive design practices,established over many years and with a sound technologicalbasis, need not be thrown out to make way for new ways ofthinking. Instead, we must look to build on prescriptivemethods and, with the use of rapidly evolving computationaltools, begin to explore more fully the world of performance-based design in engineering. Examples abound inperformance-based engineering design, many making use ofthe most sophisticated tools available today, such ascomputational fluid dynamics (CFD) modeling to computer-aided manufacturing processes. These are discussed in moredetail later, but first we need to establish more directly thedistinction between performance and prescriptiveapproaches to engineering design.

PRESCRIPTIVE DESIGN AND OPTIMIZATION VERSUSPERFORMANCE-BASED APPROACHA prescriptive approach is the most commonly adoptedapproach in engineering design. Prescription implies thatthere is a set of rules that need to be followed, rulesnormally outlined by a code or a design guide that is basedon previously developed empirical and scientific knowledge.A simple example is the design of a beam from reinforcedconcrete. The design of reinforced concrete structures, priorto the early 1900s, required a leap of faith or a belief in themany opinions and different calculation methods available

Page 121: B. Kolarevic-Performative Architecture

114

at the time. Today, the design of a concrete beam followsstandard rules of practice, so much so that we can lookto tables for most element sizing. This is not to be seenas detrimental as it allows us, as building engineers, toprocess certain aspects of design quickly. We arestanding on the shoulders of those who have come beforeus, and using this knowledge is important to speeding upthe process of design.

But prescriptive design takes on larger importancewhen one views it with respect to the larger scale designof systems, such as structural or mechanical systems forbuildings. Often due to economic and time pressures,building form is limited in complexity in order tosimplify large-scale inputs for engineers —simplifications that manifest themselves in ensuring thatthe design and the spatial configuration have been seenbefore and that the design process is predictable. Lessonslearned or technologies developed from the previousdesign of a similar building type are utilized to producequick and normally economic designs. The design processcan thus be a linear one with the rest of the design andconstruction team, since the design outputs are alreadywell understood at the outset and risk is minimized. Incommercial practice, this approach is dominant as it ismore time efficient and less encumbered by the unknown.For much of the building practice, the predictability ofknown inputs and outputs is welcome.

It is important to recognize that the engineeringefforts in this type of prescriptive design work areconcentrated mainly in the optimization of the system.Design evolution can often be categorized in quantities,such as pounds of steel or tons of cooling — traditionalmarkers of “efficiency.” Commercially availableengineering design tools today are predominantly aimedat this optimization process.

Engineering design is often hijacked by theoptimization process, and while this can be an importantpart of engineering design, it should not be misconstruedas more significant advances or quantum creative leapsin engineering design — advances that are in large partdue to thinking outside of the established solution and itsassociated parameters. Also, the quantities being

optimized, such as pounds of steel for structural systems orcubic feet per minute (CFM) of air supply for HVAC systems,can often start to change as market conditions evolve,leaving established and previously optimized solutionsbehind the curve.

It is inevitable that business and the human conditionscombine at certain points to challenge past performanceand to question previously held convictions of how toapproach a design. A prescriptive approach may providereduced risk, but it can also lead to reduced gains. Aperformance-based approach may be more tailored to usein a particular project — one where the design problemcannot be simply categorized, or a solution from the past bereadily adopted. Performance-based design offers a processthat relies more fully on an engineer’s training in creativeproblem-solving and applying first principles for design. Itdoes not preclude the use of prescriptive areas of designwhere they may be applicable, either at a component orsystems level. But in a performance-based design process,the design inputs have to be carefully developed andmeticulously understood. Their effect on the design iscritical to the development of an innovative product. Theprocess of design feedback is also critical to the success inperformance design. Finally, a design output cannot beprejudged or biased. The process must be relied upon toproperly assess inputs to develop and shape an unknownoutcome. In this way, emerging computational tools arecritical to the process, not as tools for optimization but astechnological guideposts for solving complex problems.

It is a fact that prescriptive processes are utilizedconstantly in building engineering, as our processes forperformance-based design are in their infancy. This is forgood reason, as performance-based design eliminatesseveral hurdles, the greatest of which is the requirement forintegrated design thinking for all members of a designteam. For performance-based design to work effectively, itrelies on well-communicated feedback loops betweendifferent members of a design team. This, in turn, creates aprocess that is more non-linear than a standard prescriptiveapproach — a design hysteresis loop than can often befrustrating and time consuming. Again, the use ofcomputational tools can assist and improve this process, but

Page 122: B. Kolarevic-Performative Architecture

115

the real challenge is in the management of the designprocess and in ensuring that design iterations concluderapidly. An integrated approach to engineering design— one that adapts quickly to other disciplines and isrooted in a fundamental understanding of a variety ofbuilding systems — is essential to the success of aperformance-based approach.

THE QUANTITATIVE BIAS TO DESIGN: EVOLVINGTOOLS FOR PERFORMANCE-BASED DESIGNIf engineering design processes are often behind thetechnological curve, today’s engineering design tools,on the contrary, are increasingly sophisticated andadvanced, opening up many opportunities well beyondestablished prescriptive design methods. There areseveral obvious reasons for this current development inthe engineering profession. Certainly one is thatbuilding and infrastructure engineering are benefitingfrom crossover technology from other disciplines, suchas aeronautical engineering, where computationalmodeling is significantly more advanced. Whilecomputer-aided design (CAD) use is widespread, the useof higher end computational tools is only now becomingmore typical in the offices of building engineers.Second, the rapid increase in computing power, coupledwith the strong and ever-developing software industry,has fueled the emergence of sophisticatedcomputational tools. The evolution of computing powerwill march on and we should be aware of what changesthis will bring about in our ability to accurately modeland simulate for building engineering.

Finally, and most importantly, is an overall quantitative biasto engineering design, which is emphasized both in academicinstitutions and by society as a whole. The bias is felt oftenin areas of “softer” engineering, or where design isemphasized over scientific development. The drive forbuilding engineers to become scientists is real, and thereliance on sophisticated computational tools is a byproductof this hidden sociological emphasis. This bias has a positiveside for the design engineer, of course, and computationaltools are used primarily today to provide “proof” ofpreviously recommended designs. However, computationalpower and crude design processes have in the past hamperedefforts to use computational tools as real design drivers. Thechallenge to today’s engineers involves seeking real ways ofmoving computational tools from simply being a means ofproving design ideas to being integral parts of the designprocess, parts that can provide design input quickly anditeratively. This is essential to bringing performance-baseddesign solutions to bear on problems. Areas of currentinterest to Buro Happold in performance-based designinclude structural engineering, advanced HVAC design, CAD/CAM (computer-aided manufacturing) processes and buildingsimulation.

PERFORMANCE-BASED DESIGN EXAMPLESIn the structural engineering field, finite element programshave been established as the primary computational tool forthe past thirty years. Recent advancements in graphic andCAD modeling software have aided the development of moreaggressive structural forms, and fortunately commerciallyavailable finite element programs have kept up the pace. Formore specialized areas of structural engineering, many firmshave developed performance-based software suited forspecialized uses. Buro Happold developed a non-linear finiteelement program, called TENSYL, for the original purpose ofanalyzing cable nets and tensile structures. Pure tensilesurface structures are a good example of performance-basedstructural design (figures 8.1 and 8.2). The forms for thesestructures are not designed, in as much as they are formfound from boundary conditions with surfaces calculatedbetween these boundary inputs. The surfaces themselves arenon-geometrical, as they do not follow any mathematical

8.1Tensile fabricstructure roof,exterior view. Non-linear finite elementprograms allow formdevelopment and theanalysis of non-geometrical shapes.Darien Lake Pavilion,Buffalo, New York(1996), architect FTLDesign EngineeringStudio.

8.2Darien Lake Pavilion:tensile fabric structureroof, internal view.Structures are createdfrom patterned fabricpanels into large three-dimensional surfaces.

Page 123: B. Kolarevic-Performative Architecture

116

formulae. Structures such as this have only beeneconomically achievable with the assistance ofcomputational models for load analysis and surfacepatterning for fabrication (figure 8.3).

More recently, TENSYL has evolved into a powerfultool for amorphous structural and geometrical analysis, asshown in its use on the Stuttgart 21 project (2001–) whereintricate physical models constructed by the architecturalteam were reconstructed computationally for the firststage of analysis for the project (figures 8.4–8.6). In thisprocess, tensile mesh models were form-found to developcompressive shell surfaces that make up the roof of a newmajor train station in Stuttgart, Germany. The shellsurfaces were developed with openings to allow lightpenetration from the upper level park down into the newtrack levels.

Other recent efforts to investigate economicalconstruction of amorphous forms have used performance-based design. The use of principal curvatures is a methodto economically “map” a curving surface into flat planes,avoiding the use of costly curved glass or other component-cladding materials. Buro Happold again developed aspecific software routine to use this method effectively on arange of surface structure projects. One example is thedesign of a recent 20,000 square foot glass canopy project(Roppongi Canopy, Tokyo, Japan, 2004–; figures 8.7 and8.8). In this example, the structural system was designedto facilitate the principal curvature lines for achieving flatpanels over the curving structure surface.

8.3Darien LakePavilion: computer-aided fabrication ofpatterned fabric roofpanels.

8.4Stuttgart 21,Stuttgart, Germany(2001–), architectChristoph IngenhovenOverdiek ArchitektenGmbH und Co. Newrailway station roofmade of concreteshells formed withtensile membranemodels.

8.5Stuttgart 21:computer renderingshowing staggeredindividual shellunits.

8.6Stuttgart 21:physical models forthe tensile formswere recreatedcomputationallyprior to “hardening”into stiff shellelements forstructural analysis.

Page 124: B. Kolarevic-Performative Architecture

117

Materials offer some of the most exciting advances inperformance-based design, as design processes andprojects can be developed specifically to suit the material’scharacteristics. The use of standard paper or cardboard inconstruction has been used by Buro Happold on severalprojects, including the Japanese Pavilion in Hanover,Germany (2000). While it has compressive strength alongthe lines of wood, it is limited in application by its overallstiffness and a propensity to creep with time. However, as acompletely recyclable material and one availableinexpensively throughout the world as spiral wound tubes,it is an attractive material for temporary structures. Aperformance-based approach was necessary as thecardboard material has no established prescriptive designguidelines. Through extensive testing of the strength andstiffness limits of the material (figures 8.9–8.11), BuroHappold was able to develop suitable analysis and designprocedures for the material and have the confidence tobuild with it on a large scale.

8.7 (left and right)Roppongi Canopy, Tokyo,Japan (2004–), architectSkidmore Owings andMerrill (SOM). The initialmodel where straight linesare laid on a curvingsurface resulting incurved glass claddingpanels throughout.

8.8 (left and right)Roppongi Canopy: thedeveloped surfacemodel with principalcurvatures providingthe continuous curvingsurface with flattrapezoidal glasspanels.

8.9Japanese Pavilion,Hanover, Germany(2000), architectShigeru BanArchitects.Exterior viewshowing cardboardgridshell underconstruction.

8.10 (right)Japanese Pavilion:interior finished spacewith paper claddingand cardboard tubestructural system.

8.11 (far right)Japanese Pavilion: theextensive testing ofcardboard tubes wasrequired in order toestablish designguidelines for the useof the material inconstruction.

Page 125: B. Kolarevic-Performative Architecture

118

8.12The Experimental Media and Performing ArtsCenter, Rensselaer Polytechnic Institute, Troy,USA (2003–), architect Nicholas Grimshawand Partners. The CFD model of the inside ofthe concert hall, showing relative air velocitiesin the displacement ventilation system.

8.13The University of Michigan BiomedicalResearch Building, Ann Arbor, USA (2001–),architect Polshek Partnership Architects. Theview of the CFD model for the double façadecladding wall.

8.14The University ofMichigan BiomedicalResearch Building:various CFD modelsfor lobby winter andsummer environmentalmodeling.

Page 126: B. Kolarevic-Performative Architecture

119

Mechanical engineering and, in particular, HVACdesign have long been resistant to change and havebeen slow to adopt computational tools. Significantadvances in recent years have made CFD modeling animportant part of building engineering, in particularthe development of HVAC systems for spaces whereconventional, prescriptive approaches are notappropriate. Figure 8.12 shows a CFD model of theinside of a concert hall for the new RensselaerPolytechnic Institute’s Experimental Media andPerforming Arts Center (Troy, USA, 2003–). Themodel was required due to the proposed displacementventilation solution, where low velocity air isintroduced through plenums below the seating areas.The CFD models were used to refine the HVACstrategy and to ensure that the internal comfort wasmaximized for the system. Other examples include alarge lobby space and double façade cavity wallsystem for a biomedical research facility located atthe University of Michigan (Ann Arbor, USA, 2001–).CFD modeling was used extensively to underpin theenvironmental strategies put forward by thearchitectural design team for the project (figures8.13 and 8.14).

In addition to CFD, energy modeling and thermal and lightingsimulation are quickly becoming essential for low energy andhigh-performance building system design. Environmentaleffects, at both a micro and macro level, can now be testedand are able to shape and inform critical design decisions.Figure 8.15 shows pollution dispersion modeling for the BBCWhite City project, London (2002–). The CFD simulationcritically reviewed the pollution generated by the campuspower generation equipment. Additional simulation modeling(figure 8.16) shows the effect of light pollution on the site.

CAM tools have also been long associated with advancedfabrication solutions to building components. Again, recentadvances, primarily in architectural modeling tools and theproposals of aggressive and non-geometrical forms, havequickly been pushing CAD/CAM technologies forward. A newgeneration of architects is now becoming familiar with rapidprototyping, 3D printing and a whole host of new CAD/CAMtechniques that were traditionally developed for industrialdesign applications, primarily in the auto industry. Thesetools affect not only the product but the mindset of the newdesigners, and it is clear that CAD/CAM processes for large-scale buildings are gaining significant momentum.Engineering designers who are able to not only use these newsoftware programs but, more importantly, use these

8.15BBC White City,London (2002–),architect Allies andMorrison. The CFDmodel of external airpollution on site.

8.16BBC White City: thecomputational modelindicating levels oflight pollution.

Page 127: B. Kolarevic-Performative Architecture

120

fabrication processes to inform their designs, will push theboundaries of performance-based building design in thefuture. For example, figures 8.17–8.19 show the GatesheadMusic Centre’s roof (2003), a structure developedparametrically to maximize the use of consistently radialprimary and secondary steel framing. The British Museum’sroof (2001) also illustrates an example where CAM madethe project economical and achievable (figures 8.20–8.23).

Finally, simulation tools involving human behavior arenow becoming useful in developing and testing new designstrategies for public spaces. People movement softwaresimulates individuals and response times to either life safetyevents or simply the morning rush hour: these are thencalculated with respect to the building design and layout.Figures 8.24 and 8.25 show the use of people flow modelingfor crowd control around the new Arsenal Stadium project(2001–). These simulation tools provide a great leapforward into the next generation of performance-baseddesign.

What may be most encouraging about today’sengineering design tools is that we are only at version 1.0.Computational power, in accordance with Moore’s law, willcontinue to drive what engineers can do and in what timeframe. Shorter periods for computational modeling willcontinue to allow more design feedback loops and willhopefully lead to more integrated design for building andinfrastructural teams.

8.17Gateshead MusicCentre, Gateshead,UK (2003), architectFoster and Partners.The roof structureprovides clear spancover for the musiccenter. GatesheadMusic Canopy,Gateshead, UK(2003), architectFoster and Partners.

8.18Gateshead MusicCentre’s roof: steelrib structure underconstruction.

8.19Gateshead MusicCentre: parametricmodeling for roofstructure to provideeconomy inconstant circularradius primary andsecondarymembers.

Page 128: B. Kolarevic-Performative Architecture

121

8.20Great Court, BritishMuseum, London(2001), architectFoster and Partners.Interior viewshowing the curvingglass roof over thecentral courtyard.

8.21The BritishMuseum:overhead view.

8.22British Museum:drawing of differingelement sizes forroof structure.

8.23The British Museum:the CAM of all nodesallowed a largenumber of preciseconnections to befabricatedeconomically.

Page 129: B. Kolarevic-Performative Architecture

122

SUMMARYComputational tools are clearly gaining ground in avariety of uses and their practical applications in manyareas of building engineering design are helping topush the boundaries of quality and complexity inarchitecture and infrastructure. The challenge in theeffective use of these tools will be how well they areadapted to specific problems, and how well they areintegrated into the design process, a process that todaycan be free to be more performance-oriented than amore traditional prescriptive approach.

But let us not forget that the best computationaltools do not offer solutions for design. They merely hintat this, and it is the responsibility of the designer toensure quality and coherence in the design process.

Finally, it should be evident that today’sperformance-based approach may well be tomorrow’sprescriptive recipe for a solution. The only constant inour world is change. And as engineers we need notresist this critical process of improvement of the builtenvironment, but we must seek to think outside of thebox for solutions for tomorrow’s problems — solutionsthat ultimately ensure the highest design quality forour clients.

8.25Arsenal Stadium:crowd flow modelingindicating congestionpoints created by thelarge exodus ofpedestrians after agame.

8.24Arsenal Stadium,London (2001–),architect HOK Sport.People flow modelingis increasingly beingused for assessingsafety and comfortconcerns at publicvenues.

Page 130: B. Kolarevic-Performative Architecture

123

9

JEAN-FRANCOIS BLASSEL

ENGINEERING IN

ENVIRONMENTA PERFORMATIVE

Page 131: B. Kolarevic-Performative Architecture

124

9

JEAN-FRANCOIS BLASSEL

ENGINEERING IN

ENVIRONMENTA PERFORMATIVE

Page 132: B. Kolarevic-Performative Architecture

125

RFR’s work on architectural structures and buildingenvelopes has positioned the firm at a threshold thathas long separated architecture and technology. Thatposition has fueled the firm’s interest in detecting apotential convergence between the two disciplines.Could it come now from the emergence of new modesof practice in both fields?

The practice of engineering has been moving fromconcepts and ideas towards concrete objects. Today,however, the analytical process tends to be displacedby modeling. Complex simulations of completeconstruction are increasingly being used to replacethe rational process of unfolding a situation or aproblem into a collection of individual, simple partsfor eventual recomposition.

The mastery of the analysis is being displaced bythe definition of the criteria — architectural as wellas technical — that will segregate a successfulexperiment from an unsuccessful one. The criteria arenot always quantifiable but all can be considered aspart of an economy — monetary (always a powerfuldeterminant in engineering) but also an economy ofmaterials, of means, and even an economy of time.

CHANGES AND CONVERGENCESThe work of RFR revolves around architecturalstructures and building envelopes. We often findourselves operating in a fuzzy zone where architectureand engineering overlap. This overlap explores apotential convergence between the two disciplines,and examines the ways in which the work of architectsand engineers is being transformed by this proximity.

Architects who understand and accept theformidable power and responsibility that comes withmodern technology have a great advantage inestablishing their own methods for designing buildingsfor today — buildings that respond ethically to thecurrent situation — sustainable, graceful, solid andeconomical. Very few teams have, so far, managed totranscribe this understanding into their practice. It isnot only the architect, however, whose role is beingtransformed. Engineers, the traditional bearers of

technology to the design process, also see their roleprofoundly transformed. Individual engineering specialtiescannot work neatly in isolation anymore, ignoring eachother. Engineers, as much designers as the architects,develop a capacity to work collectively and to take part ina project laterally. They must be not only highlycompetent in their often narrow field but also be able toreplace their problems in a more global vision of theproject or of the environment. Thinking about a project asa continuous transformation of the environment requiresspecialization and openness, invention and integration,and no blind spots. The emergence of a type of practicebased on integration will hopefully liquidate the measureof technological performance by isolated quantifiablecriteria. The criteria that spring from this new situationare much less definable and the engineering morecomplex. Borders are blurred: engineers tackle questionsof architecture and offer solutions to make architecture.

Engineering has been described as an analyticalprocess where one moves from concepts to a technicalsolution. This abstract process of analysis is being rapidlydisplaced by simulation, profoundly changing the way wework. Problem solving by analytical processes is beingreplaced by more complex simulations of complete builtenvironments, from structure to cities. Of course, no toolis capable of an exhaustive simulation — but the masteryof the analysis is being displaced by the definition of thecriteria. Choosing the right set of criteria by which toappraise the simulated projects becomes the core of thedesign: you design but at the same time you try tounderstand that which you are designing. These criteria,architectural as well as technical, create a dichotomy:they separate the successful experiments or models fromthe unsuccessful ones.

Changes can also be felt in the influence of economyon buildings. Monetary economy has always been apowerful determinant in engineering but there havealways been other forms of economy, measured indifferent terms: there is an economy of materials, aneconomy of means, and an economy of time, for instance.If the nineteenth century was concerned with the economyof materials and matter, and the twentieth century could

Page 133: B. Kolarevic-Performative Architecture

126

be best described by the search for an economy oflabor and ultimately an economy of time, then theeconomy of the building industry in the twenty-firstcentury will be an economy of energy, in all its forms— not only the drain on resources induced by thedesign during its useful life but also what itembodied at construction and, of course, whathappens after the design is no longer useful. Thesetwo aspects of the changing conditions in which wedesign seem to influence the way in which buildingsare made on at least two levels — where technologyand architecture converge.

Technology influences, more directly than ever,the external morphology of buildings: directly, sincethe building’s skin is the locus of exchanges ofmatter and energy between the building and itsenvironment, and indirectly, since modeling toolsallow almost surreal explorations of form. Morefundamentally, buildings are technical objects —extensions of our bodies that allow us to do a numberof things that we cannot do naturally.

Buildings are evolving in the manner of otherclasses of technical objects. The research of theevolution of the modern technical object generallyreveals, at its genesis, a mode of composition basedon juxtaposition and addition — the joining ofautonomous parts; then, as technical objects becomemore sophisticated, their components tend tocondense. In the case of the automobile, for example,the engine that was once sitting passively in thechassis plays a role in the rigidity of the frame oftoday’s car. It is no longer independent from thebody or the windshield. When applied to buildings,this design philosophy is manifested by the activationof building components. Elements or components ofa building play multiple and complementary roles inorder to increase the building component’s physicalusefulness beyond its initial destination.

The four projects that follow, one completed andthe other three work in progress, illustrate theseintroductory remarks.

AVIGNON TGV TRAIN STATIONThe TGV1 train station in Avignon, France, opened in 2001.A multiplicity of criteria had to be considered andarbitrated during its design with the in-house architecturaldepartment of the SNCF, the French railway authority.

The station sits at the confluence between the Rhoneand Durance rivers. It is a flat, hot, and windy site in thesouth of France. A very strong wind blows almost one dayout of four. High rows of cypress and poplar trees protectthe houses and fields in the surrounding area.

The elevated track and platforms increase the exposureof waiting passengers to the harsh natural elements, andthe station is designed first as a shelter from the sun andwind. It works like an airport boarding lounge where onewaits for the approaching train behind glass. The tracksideof the building faces roughly north and the entrance sidefaces south. Despite its symmetry, the building sectionreflects the differences, of climate and use, between the twosides. Less space is needed at the end than at the center ofthe building, so the building tapers towards the end.

A large number of parametric studies, looking for abalance between repetition and tapering form, led us tochoose a rotationally repetitive geometry. The buildingspace is bounded by a small volume resulting from theintersection of two tori (figure 9.1), surfaces more easilydescribed than the curves they generate where they meet.The two tori, each with its set of structural axes, generatetheir own structure (figure 9.2). The different qualities ofthe two opposite skins inform not only the surface but alsothe depth of the structure. Although it is shaped like agothic arch (figure 9.3), the climatic asymmetry of the skintells the structure what it wants to be. On the south side,the opaque skin shields the building against the sun and isvery thick and stiff, although relatively light because of thehigh risk of earthquakes. Its construction relies on anunconventional use of conventional techniques: open sectionsteel work, steel decking, insulation, waterproofing andcladding with a glass cement composite panel. The otherside is totally different. It is as much driven bytransparency as the opposite side is driven by opacity. Lightand views are key.

Page 134: B. Kolarevic-Performative Architecture

127

The architecture of the insulating glass panels,deduced from the ridgeline for optical reasons, losestwo key benefits offered by the underlying toroidalgeometry: flatness and repetition. Instead, the panelsare warped and unique. Their supporting structureseeks its slenderness — and transparency — in theadjacent building elements. It borrows its rigidityfrom the structure of the south side and from theconcrete sub-level. The glass composition combinesmetallic coatings and ceramic frit of varying densityto achieve optimal thermal performance as well asmaximum transparency (figure 9.4).

The search for transparency led to areconsideration of thermal standards. Contemporaryenergy codes severely restrict the amount of glass in abuilding envelope. Their strict application would haveconsiderably hindered the design. However, in a trainstation, where people wear coats in winter, the interiortemperature can drop to 14°C without creatingdiscomfort. The building complies with energy codesdirectly and not by the installation of a standardizedpredetermined amount of insulation.

Optimal building performance comes fromlooking at things in a certain way. One gleans most ofthe necessary information from the singularities of thesite and the future use, and not from some abstractand generic way of thinking. It is very difficult toattach a single criterion to a particular piece orcomplex of pieces, be it structure, skin or any of thebuilding elements. They all interact with each otherand one cannot select and decide to focus only on oneaspect of performance to determine what the designof the building should be.

9.1The toroidalgeometry of theTGV train stationin Avignon, France(1998–2001),architect J. M.Duthilleul,engineers RFRConsultingEngineers, Paris.

9.2TGV station:structure underconstruction.

9.3TGV station:typical section.

Page 135: B. Kolarevic-Performative Architecture

128

9.4Interior viewof the trainstation inAvignon.

Page 136: B. Kolarevic-Performative Architecture

129

BERCY FOOTBRIDGEThe design of a bridge is very heavily influenced byquestions of spanning, but even with very long spans,other things inflect the design. To say that a structure isexclusively dictated by structural rationality ismisleading.

The Bercy footbridge2 spans the River Seine betweenthe Park of Bercy and the French National Library inParis. The structure is made to cross the river, but it isalso made to link places in the city so that the paths inthe city inform the way the structure is built (figure 9.5).Rather than building a span and then finding routes atthe end, structure and path merge. Here, the bridge itselfprovides paths, not only across the water but alsobetween the low banks of the river and the high level ofthe esplanade of the bibliothèque on the Left Bank andthe park on the Right Bank (figure 9.6).

There are several ways of looking at the structure,but one of them is simply to look at the oppositecurvatures of the paths as the top and bottom membersof a beam with two hinges (figure 9.7) — a lens-shapedtruss in the center and cantilevers at each abutment(figures 9.8 and 9.9). The public path is also what givesthe structure its rigidity.

9.5Bercy footbridge spanning between the Parkof Bercy and the French National Library inParis (1998–2005), architect FeichtingerArchitects, engineer RFR ConsultingEngineers.

9.6Bercy footbridge: a compositecomputer rendering showingthe bridge’s multiplepathways.

9.7Bercy footbridge: theopposite curvatures ofthe bridge’s pathways.

9.8Bercy footbridge:finite elementsanalysis of structuralstresses.

9.9a–bBercy footbridge:simulations ofstructuralbehavior.

Page 137: B. Kolarevic-Performative Architecture

130

BANDSHELLThe next example is a very small project, only acouple of hundred square meters: a bandshell for anopen-air piano festival held in an old park inProvence.3 Despite its modest size and apparentlysimple brief, a great deal of interaction betweenvarious factors occurs when seeking a single shapethat would be equally satisfactory from the point ofview of acoustics, structure and form: anacoustically efficient light garden structure (figure9.10). The shell was first seen through an opticalanalogy, leading to an initial geometry. Modeled andexplored through software that simulates theintelligibility of sound, the initial geometry yielded afirst revision, itself the starting point for a set ofanalyses looking at the structural behavior of the

shell. Increasingly detailed models helpedunderstand how to adjust the geometry, global andstructural, in a way that would remain compatiblewith the acoustic and structural behavior of theshell.

This type of exercise on such a low budget,small-scale structure has become possible onlyrecently thanks to inexpensive, user-friendly, reliablesoftware. Significant to what is happening in botharchitecture and engineering is that the tools changenot only the quantity of things we can do but alsothe quality. Different types of investigations canhappen much earlier in the design process. A greatdeal of interactivity, not previously possible, can nowtake place realistically. The change of paradigm hasjust begun.

9.10An acousticallyefficient lightstructure for abandshell in a park,Provence, France(2003), architectExplorations,engineer RFRConsultingEngineers.

Page 138: B. Kolarevic-Performative Architecture

131

STRASBOURG TRAIN STATION EXTENSIONThe last project is an extension of the old trainstation in Strasbourg, again with the help of theSNCF’s in-house architectural department and ofMatthias Schuller’s team in Stuttgart. Trains stopone level above the access level, while publictransportation runs 15 m below ground. A link isrequired in front of the station, between the streetlevel and the underground tram. However, thestation is one of the urban features of the city andthe new building extension should not hide it.Therefore the project seeks to maintain thehistorical presence of the train station, to link thevarious points of access to the various modes oftransportation and to provide proper shelter in allpossible climates.

The initial answer, a glass bubble along the south sideof the station (figure 9.11), had to be considerablyrefined in view of the first thermal simulations. Theskin of the building, a filter for solar energy and heat,dominates the design so much that the work focusedfirst on resolving questions related to the behavior ofthe skin before tackling its supporting structure. Twoquestions arose. First, how can the geometry beresolved rationally to match objectives of space,shape and form? Generations by translation androtation are being tested now. Second, what strategywould provide climatic comfort with a minimumexpenditure of energy?

Various simulations have shown that only acombination of elements, adjusted to their orientationon the surface of the bubble, will let the skin shed

9.11The proposedextension of thetrain station inStrasbourg,France (2003–07), architectand engineerRFR ConsultingEngineers.

Page 139: B. Kolarevic-Performative Architecture

132

most of the incoming energy. The weather barrier ismade of a glass combining a good heat mirror,ceramic frit screening, low-emission coatings, andsomething similar to a temporary double skin. Alow-emission curtain drawn inside the glass bubblecontains the heat at the surface of the glass andconducts it to the top of the building where itescapes through openings.

The most interesting discovery of this projectwas the unexpected opportunity offered by thetunnel dug under the station. Strasbourg is builtalong the River Rhine and the tunnel bathes in itswaters, effectively turning it into a giant cooling

system (figure 9.12). By carefully designing thebuilding to help pull in more air than would come outnaturally from the tunnel (and by adding a couple ofsimple, low-energy devices like a radiant slab, itselfdrawing from the water table), it is possible to makethe space under this very large south facing glass wallcomfortable in summer, almost exclusively throughpassive means. Now that the composition of the skinis known, we will design a structure that works withit. Again, the nature of the skin drives theorganization of the structure and the nature of theskin is determined as much by optical criteria as bythermal ones.

9.12Section through theproposed extensionof the Strasbourgtrain station.

Page 140: B. Kolarevic-Performative Architecture

133

CONCLUSIONSThis is not really a conclusion but a reaction to whatwas described earlier. Can we really separate architectsfrom engineers by what can be described by numbers?Our culture, every single human culture in fact, can bedefined by its technology. Why would one want toexclude technology from any discussion or be afraid ofrecognizing technology as part of our culture, to beunderstood, appreciated and judged? Ultimately it hasto do with democracy — one cannot leave technology tothe market. Architects have a responsibility to usetechnology well and not hide it — or hide from it —and bring it to the public.

NOTES1 TGV – Trains de Grand Vitesse — France’s high-speedtrain network.2 The Bercy footbridge was designed by Henry Bardsley.3 A competition won with Explorations, who are formerstudents of ours.

Page 141: B. Kolarevic-Performative Architecture

134

Page 142: B. Kolarevic-Performative Architecture

135

10

HARALD KLOFT

NON-STANDARDARCHITECTURE

NON-STANDARD

DESIGN FORSTRUCTURAL

Page 143: B. Kolarevic-Performative Architecture

136

10

HARALD KLOFT

NON-STANDARDARCHITECTURE

NON-STANDARD

DESIGN FORSTRUCTURAL

Page 144: B. Kolarevic-Performative Architecture

137

This chapter examines structural engineering in aperformance-based architectural design environment,presented against the background of digital workflowprocesses of developing and producing the so-called“free-form” or “non-standard” architectures.Referring to the architecture of complex geometries as“freeform” is somewhat controversial because theterm tends to simplify the conceptual and intellectualcomplexity of recently developed digitally-driventechniques of form generation. The curators of arecent exhibition1 at Centre Pompidou in Paris insteadused the term “non-standard architecture” to refer tothe architectural objects expressing the new formalfreedom. Whether “freeform” or “non-standard,” anumber of recently completed projects with complexgeometries manifest a move away from traditionalformal principles in architecture, often demandingnovel approaches to structural design and engineering,in which performance-based design and digitalworkflow are closely interwoven.

What is “free” in a freeform? If one compares thebiomorphic structures found in nature with thematerial realities of tectonic languages in building, anobvious difference is that there are no straight linesand rectangular geometries in nature. This is becausethe form generation processes in nature are based on

the rules of evolution and are optimized for adaptability inthe natural environment. In contrast, formal and tectoniclanguages in architecture are decided mostly by thecapacities of the technical production, i.e. by availabletechnologies in a given age. With industrialization, thetechnical possibilities of production increased substantially,making departures from basic geometries possible inarchitecture. The formal, geometric shifts often coincidedwith the development of new techniques and materials. Thiscorrelation is historically obvious; for example, in the1950s, 1960s and early 1970s, developments in concreteand later plastics and textiles inspired architects andengineers to move away from Pythagorean geometries andto treat form in a less restrained geometric manner. Well-known examples are the filmy shell structures of FelixCandela, the structurally logical ornamentation of thePalazetto dello Sport in Rome built in 1957 by Pier LuigiNervi2 (figure 10.1), Eero Saarinen’s concrete sculpturalforms for the TWA Terminal in New York, completed in1962, or Matti Suuronen’s utopian plastic vision for“tomorrow’s living” called the Futuro House. All theseexamples manifest the technical potential of this epoch andare expressions of the spirit of the time, like Frei Otto’slightweight tent- and cable-net structures which reachedtheir zenith in the Munich Stadium for the Olympic Gamesin 1972 (figure 10.2).3

10.1The curved ribs of thePalazetto dello Sport inRome, Italy (1956–57),architect and engineerPier Luigi Nervi.

10.2The Stadium for the 1972Olympic Games, Munich,Germany (1968–72),architect Günter Behnisch,form-finding Frei Otto.

Page 145: B. Kolarevic-Performative Architecture

138

It is important to note that the formal “freedom” inthat age (1950 to early 1970s) was limited because ofthe lack of high-performance hardware and softwaretools (“tools” in the sense of computer programcomponents) in design development andmanufacturing. Forms were often found in experimentswith scaled physical models, as manifest in the work ofFrei Otto or Heinz Isler. The forms they designedthrough a model-based form finding process werestructurally optimized by following the rules ofphysics.4 They stand in sharp contrast to the formsdesigned by form generation processes inspired by non-technical issues, for example in the work of FrederickKiesler (who began his architectural formal researchbefore World War II) and later on in the utopian ideasof Archigram and others.

Finding a structurally optimized and geometricallyclearly defined form was a necessary condition forbuilding, i.e. for material realization, in the pre-digitalera. Frederick Kiesler, however, was not interested indefining forms in a geometrically exact manner thatfollows physical logic. For his design of “The EndlessHouse” (figure 10.3), Kiesler made numerous freehandsketches to visualize his ideas about the form. Hisnaturalistic design was celebrated as the “biomorphicanswer and antithesis of the cubistic architecture of

modernists.”5 For Kiesler, form does not follow function:form follows vision and vision follows reality. Tocommunicate the spatial complexity of his ideas, he wouldcreate physical models just as a sculptor would model anart piece. Unlike the projects by Frei Otto and Heinz Isler,the form of Kiesler’s “Endless House” was not inspired bystructural optimization but by careful proportioningdriven by the scale of human beings in the naturalenvironment.6

ARCHITECTURAL FORM GENERATIONRecent years have brought a renaissance of freeforms —designers today are formally less constrained as a result ofadvances in computer technology. In Kiesler’s form-generation process, the sketches and physical models werenot linked — they had the same origin but the geometrieswere not exactly the same. Today, a digital model thatlinks design and manufacturing processes is a basiccondition for realizing freeform architectures. How thedigital models are generated depends on the designprocess, which can vary distinctly from one designer toanother or even from project to project. In addition, eachgeometrically complex project presents differentchallenges for structural design and can result in differentdigital working processes, as each of the four projects7

described below demonstrates.

10.3Endless House(1961), architectFrederick Kiesler.

10.4Dynaform BMW Pavilion, 2001International Motor Show,Frankfurt, Germany (2001),architects Bernhard Franken andABB Architekten.

Page 146: B. Kolarevic-Performative Architecture

139

DynaformDynaform is the name given to the BMW pavilion projectdesigned by Bernhard Franken (in association with ABBArchitekten) for the 2001 International Motor Show inFrankfurt (figure 10.4). Bernhard Franken and his team usecomputational tools to generate architectural form; there isno preconceived formal idea at the beginning of a project. The“parametric design process”8 of the Dynaform started with abriefing by the client about the communication message forthe exhibition; BMW’s primary interest was the presentationof the new “7” series. Franken translated program- and site-specific parameters into virtual forces using Maya animationsoftware.9 He set up a three-dimensional (3D) matrix that wasinitially shaped according to the virtual forces of a driving car(the influence of the program). Adjacent buildings on the site,such as Nicholas Grimshaw’s Exhibition Hall, had additionalimpact on the shape through a series of specially designedforce fields (figure 10.5). Through a time-based (4D)modeling process, the initial shape was deformed and alteredby the software, until an architectural form was found bysampling the generative process (figure 10.6). Theapproximate shape of the found form was corrected forgeometrical errors and established the 3D “master geometry”of the project (figure 10.7). That “master geometry” providedthe dimensional reference for all project participants duringdesign development and construction.

10.5Dynaform: theforce-field setup forthe form generationprocess.

10.6Dynaform:sampling of theform generationprocess.

10.7Dynaform:the “mastergeometry.”

Page 147: B. Kolarevic-Performative Architecture

140

M.art.A. MuseumFrank O. Gehry’s M.art.A. Museum (Museum for Furniture,Culture and Fine Arts) in Herford, Germany (2000–04; figure10.8), illustrates a different approach. Here the design processdid not start in a virtual design environment; instead, thearchitects manually built a series of physical models, many ofwhich were three-dimensionally digitized in the computer-aideddesign (CAD) environment of CATIA10 to correct and check theshape with respect to the program and the site. Gehry generallyconcentrates on the effects of the exterior surface and theinterior spaces, and relies largely on physical models to verifythat the original design intent is met. The CAD-corrected data(figure 10.9) enabled the building of more accurate physicalmodels; partial and more detailed models are also oftenproduced to explore the implementation of the overall shapingstrategy in projects. Unlike Franken, Gehry did not define a 3D“master geometry” in this project as a dimensional referencebefore starting the design and structural development. Gehry’sform-finding is an iterative process in which form changes areprogrammatically driven.

Kunsthaus GrazPeter Cook’s and Colin Fournier’s design for the KunsthausGraz, Austria (2000–03; figure 10.10), illustrates the thirdapproach. Here, the conceptual design phase (during thecompetition) did not rely heavily on computers and the physicalmodel that represented the final shape was handmade (figure10.11).11 The first impulse was to perform a 3D scan of thecompetition model in order to produce an initial 3D digitalmodel. But after considering how to optimize the form in termsof structure and materials, the architects and engineers jointlydecided to generate a new 3D digital model from scratch using

10.8M.art.A. Museum,Herford, Germany(2000–2004),architect Frank O.Gehry andAssociates.

10.9M.art.A. Museum:3D digital model ofthe building’sgeometry.

Page 148: B. Kolarevic-Performative Architecture

141

Rhinoceros 3D modeling software. While attempting toclosely follow the shape proposed initially, the digital 3Dmodel was built independently, i.e. it did not containdigitized data taken directly from the physical models,which is a process typically used in Gehry’s office, aspreviously discussed. The digital design model (figure10.12) was shaped to capture the design intent of theoriginal scheme and allowed us to optimize the form duringthe design process with regards to the structural behavior,such as its geometrical stiffness, and to address some of themanufacturing issues. This is an important difference fromthe previously described form-generation processes used byBernhard Franken, where the digitally-generated form wasdefined as a “master geometry,” meaning that later in thestructural design it was not possible to optimize the formgeometrically, i.e. by changing the shape.

STRUCTURAL DESIGNStructural design means moving from geometrical form toa structural system, i.e. connecting the formalarchitectural idea with the rules of the stress flow. Whatthen are the implications of the different designapproaches, discussed so far, for structural design? In allcases, it is important to note that the analytical softwaretools used by the engineers and the 3D design environmentsused by architects do not normally share a common digitaldatabase structure. Thus, it is in the interest of theengineers to develop data “post-processing methods” thatautomate and accelerate the geometrical data transfer.12

Inputting geometry manually is impractical and time-consuming for complexly shaped structures. Being able toimport data files accurately and quickly enables engineersto directly apply finite element and spatial vector programsto problems that need to be solved during the early designphases.

In Franken’s Dynaform, a single layer skin defined thegenerated form — the “master geometry.” The skin’smaterial and thickness, however, were not specified. Sincethe “master geometry” was fixed, we could not optimizethe structure geometrically through modifications of theoverall shape. High local forces or bending moments had tobe accepted because structural optimization of the overallshape would have called the underlying design approachinto question. Instead, the aim was to support the intentionof architecture to bring forces into a dynamic balance andto express the idea that the form is only a frozen instancefrom a long series of possible geometric constructs. Theresult was that the structural system had not beenoptimized in Dynaform.

10.10Kunsthaus Graz,Austria (2003),architects Peter Cookand Colin Fournier(spacelab.uk).

10.11Kunsthaus Graz:the handmadecompetitionmodel.

10.12Kunsthaus Graz:the digital designmodel.

Page 149: B. Kolarevic-Performative Architecture

142

As there was little prior knowledge of structuralperformance of freeforms, we focused on two optionsto materialize the form and develop the structure: wecould either design a system of linear or curvilinearstructural members (figure 10.13) that would havesupported a secondary and non-structural skin (figure10.14), or the skin itself would have been conceived asthe primary load-bearing system and would havebecome “skin deep”13 — a surface-structure withshell-like behavior. The decision as to which option toadopt was driven by a number of studies of thedifferent approaches; the final structural design wasdeveloped in close collaboration with the architects.The extremely short construction period was one of themain arguments to separate the structure and the skinby designing a primary load-bearing system of weldedsteel frames and a secondary pre-tensioned membranelayer.

In Gehry’s project, we were given the digitalgeometric model of the outer and inner surfaces (figure10.15); the interstitial, “in-between” space had to

accommodate the structural and all necessary mechanicalsystems. The load-bearing structure — often a series of steelframes — was hidden from the view and consequently wasalmost irrelevant architecturally. In this “undercover” role,the main part of the structural engineering was geometricoptimization, not with the aim of optimizing the geometricstiffness of the shape but to design a layout of structuralmembers in the interstitial space and to optimize theirarrangement so that they would work structurally (figure10.16). The internal and external surfaces themselves actedas enclosures without any primary load-bearing function;their geometry established boundary parameters thatgenerally could not be changed. As in Franken’s project, thesurfaces functioned as “master geometries” but these gavelittle opportunity for optimization of the structural behavior.Whereas the layering, i.e. the arrangement of the differentfunctional layers (structure as an inside or outside layer, orin-between) and therefore the perception of structure, wereopen questions in the Dynaform project, in the M.art.A.project the structural elements functioned only in onedimension — as a technical necessity in-between.

10.14Dynaform: thefinite elementanalysis of themaster geometryin the pre-designphase.

10.13Dynaform: thefinite elementanalysis of thestresses incurvilinearstructuralmembers.

10.15M.art.A Museum:the architectural(non-structural)skin.

10.16M.art.A Museum:the structural systemwas designed in theinterstitial spacebetween the givenexterior and interiorsurfaces.

Page 150: B. Kolarevic-Performative Architecture

143

Peter Cook and Colin Fournier, unlike Gehry, createdthe digital model of the geometry for the KunsthausGraz in the design development phase by translatingthe conceptual competition scheme into a digitalenvironment. While the objective was to remain closeto the original shape, Cook and Fournier did notpursue this dogmatically. Also, the structuralbehavior was allowed to have an influence on the finalgeometry.

For the complex roof of the Kunsthaus Graz, theprinciple task was to design a system of tubular steelmembers that would support an outer layer of acrylicglass panels with complex shapes (figure 10.17). Theexternal skin was designed as a series of discretelayers, each responding to a specific set of functionalrequirements. The structural layer, in response to theneed for a stiff and optimized structural system,consists of members arranged in a triangulatedpattern that was designed as a hybrid structuralsystem combining behavior of shell structures andbending systems (figure 10.18).

10.17Kunsthaus Graz:the finite elementanalysis of thecomplexly shapedskin.

10.18Kunsthaus Graz:triangulatedstructuralengineeringpattern.

10.19“Klöpp” ministrycomplex, Reykjavik,Iceland (2002–),architect FrankenArchitekten.

10.20aKlöpp: finiteelement analysisof the outerfaçade.

The latest project of Bernhard Franken — “Klöpp” — aministry complex of three buildings in Reykjavik, Iceland(2002–; figure 10.19), deals with the structural designin a very different manner than discussed previously. Itwas the first project in which Franken did not strictlydefine a “master geometry;” the geometry, especially inthe façades, was subject to change as the projectdeveloped. The generation of form was done digitally asin his earlier projects, using Maya animation software.The conceptual origin, as inscribed into the parametersof the initial form generation process, was found in thefissured surfaces of Iceland’s topography and in the stonemonument near the site, inside which, in accordance withthe Icelandic legends, elves should live. The formgeneration was highly iterative: the initial form,generated by sampling time-based processes modeled inMaya, was structurally analyzed using finite elementssoftware (figure 10.20a). After interpreting the results,we optimized the form geometrically in collaborationwith the architects, while adhering to the initial designintentions (figure 10.20b).

10.20bKlöpp: structuraloptimization of thebracing system inthe outer façadewalls.

Page 151: B. Kolarevic-Performative Architecture

144

MANUFACTURING PROCESSA lack of suitable materials and production techniques thatare usable in building are problems that anyone seeking todesign and build freeform surfaces is faced with at somepoint. Building regulations and standards that do not covernew procedures and materials, and potentially highconstruction costs, add to the difficulty of constructing acomplex shape. Since buildings are normally prototypes, wecannot apply industrial processes that may be well suitedfor the production of parts with complex shapes but requirelarge quantities in order to be economical.14

In the Dynaform project,15 after much discussion andthe evaluation of different options, the architects andengineers decided jointly to separate the primary load-bearing structure from the structural secondary skin and todesign a series of primary steel frames. Franken’s teamgenerated fifteen cross-sections as “derivatives of themaster geometry,” each at a different angle (i.e. they werenon-parallel) and each resulting in a unique shape (figure10.21). We then inscribed structural frames — the so-called “Dynaframes” — into those sections. The outeroutline of the Dynaframes is precisely offset from the“master geometry” surface, while the inner outlinerepresents a reversal of the same master shape (figure10.22). Both outlines are connected at regular intervalswith welded plates in order to work structurally as aVierendeel system. Those plates all point to the origin ofform generation — the virtual curvilinear path generatedby a car driving through the space (figure 10.5). The designof the rather strange looking Dynaframes thus originates inthe general form-finding principles of the scheme — it isnot primarily driven by a structural logic.

The steel members for the frames were cut from flatsteel plates, then bent into shape and manually weldedtogether. The contractor was faced with the challenge ofhaving to maintain tight tolerances while translating thedigital 3D data into a built shape. A full-size mock-up ofseveral structural frames was completed three monthsbefore the final erection (figure 10.23); it enabled us tostudy the required assembly time and procedures as well asidentify and resolve problems with the connections betweencertain components.

10.21Dynaform: cross-section outlines as“derivatives of themaster geometry.”

10.22Dynaform: thestructural“Dynaframes.”

10.23Dynaform: the full-scale mockup of the“Dynaframes.”

10.24Dynaform: the full-size mockup of asection of themonoaxial pre-tensioned PVCmembrane.

Page 152: B. Kolarevic-Performative Architecture

145

The exterior of the pavilion is meant to be the purestpossible representation of the “master geometry.” A majorconcern was to find a way of constructing a skin that wouldgenerate a smooth surface over the complex shape. Afterresearching different material options, a monoaxial pre-tensioned PVC membrane was selected.16 By modeling theskin as a ruled surface, we avoided any external folds, andthe membrane could be pre-stressed between the structuralframes. To prove that this new envelope concept would work,a full-size mockup was constructed and subsequentlyapproved by the design team and the client (figure 10.24).

The membrane layer was assembled by specialists withmountain-climbing skills (figure 10.25). The assemblystarted two months before the opening of the motor show.Each span between adjacent steel frames was covered withone membrane segment. The joints between membranes weresealed with an apron fabric (figure 10.26) — a compromiseas far as the overall appearance was concerned, butunavoidable with respect to the tight construction scheduleand the need for a watertight envelope.

10.25aDynaform:assembly of themembrane in themost difficult area— the transitionfrom thecurvilinear torectangularsurface.

10.25bThe Dynaformwith open joints.

10.25cDynaform: theskin fits tightly.

10.26Dynaform: thejoints betweenmembranes weresealed with anapron fabric.

Page 153: B. Kolarevic-Performative Architecture

146

The finished project looked effortless, with materialsand shapes appearing to connect almost naturally(figure 10.27). Designing and building a largefreeform structure, such as the Dynaform, requiresmore energy, time and creativity than would benecessary for a conventional structure. Deadlines, thebudget and the design intentions could not becompromised. The Dynaform pavilion remained inplace for the duration of the motor show; it was thendisassembled and is stored for future use.

CONCLUSIONSThe translation of freeform, non-standard architecturaldesigns into built structures requires the development ofnew modes of thinking from all project participants. It isessential that architects and engineers collaborate fromthe very beginning of a project. In the case of “freeform”architecture, an important aspect of this collaboration isthat the structural engineer has to “speak the language”of the architect and fully support the particular designapproach. Understanding individual design values isessential for a productive collaborative dialogue betweenengineers and architects.

10.27Dynaform: theseamless materialitywith complexgeometry.

Page 154: B. Kolarevic-Performative Architecture

147

The shown projects mark the beginning of newopportunities in building. Boldly curved shapes, like theDynaform and the Kunsthaus Graz project, were regardedas unrealizable a few years ago. Both projects are radicalin the way all participants — clients, planning partiesand contracted firms — believed in the idea of realizing avisionary architecture and took the risk of failure. Theplanning and manufacturing processes were characterizedby the necessity of developing new solutions from scratch.In this context, it was extremely helpful that Franken’sparameter-driven design method in the Dynaform projectdid not allow for geometrical alterations to thearchitectural model. Only with these experiences is itpossible to take on the challenges of the “Klöpp” project,where the geometry, especially in the façades, is subjectto change for structural optimization. A collaborativeprocess based on the mutual processing of parametricdata, as was the case in the "Klopp" project, is the mostpromising for the future, since it allows for the geometricoptimization of form in terms of its structural behavior.

In my experience, there are several “rules” forengineers and architects interested in designing andbuilding the “non-standard architectures.” First, full-scale mockups of the geometrically most complex andchallenging areas are absolutely essential. Problems thatcannot be solved in the production of the mockup will notbe solved on the construction site either. Second, the

digital model of the geometry for use in manufacturing andconstruction should be fully parametric in order to facilitatethe quick generation of the necessary data sets. In theDynaform project, for example, the specialists for structuralmembranes developed a method for generating a ruledsurface from the given geometry by splitting the membraneinto pieces which seamlessly connect the cross-sectionalframes in areas of similar curvatures; that is the principalreason why we could use the uni-axial tensioned membranewithout getting folds on the surface. Third, the control oflayering during the manufacturing process should not be leftto the construction company. This is an important new taskin the digital workflow and should be done by the designteam, either by architects or engineers. As a task, the layercontrol is analogical to the role of a compiler in thecomputer — it produces the detailed “machine” code, i.e.the data necessary for execution. In the Dynaform project,the 3D model of the geometry produced in the architecturalform generation process is one-layered — it consists ofsurfaces without material specificity, without thickness. Atthe end of the design development, a multilayered 3D modelis produced, which includes all the information necessaryfor manufacturing (figure 10.28). During manufacturing,however, that 3D model undergoes numerous changesdepending on the tolerances and specific demands ofproduction. As in compilers from the computer world, thetask of surface control is to provide each manufacturer witha specific layer of information that is needed.

In conclusion, the emerging digital design andproduction environment, combined with new materials andmodern technologies, offers unprecedented possibilities forarchitecture that strives to perform fully in a conceptual,formal, technical, financial and material sense. In such acontext, structural engineers can exploit fully the creativepotential of the discipline, not only by supporting thearchitects with the calculations, but by embracing theirideas and realizing the digital form generation possibilitiesstructurally and materially. The freeform, non-standard andperformative architectures offer a promise of newcollaborative design synergies for architects and engineers.

10.28Dynaform: the full,detailed 3D modelof the buildingdata.

Page 155: B. Kolarevic-Performative Architecture

148

NOTES1 “Architectures Non Standard,” December 10, 2003 – March 1,2004, Centre Pompidou, Paris, France (http://www.centrepompidou.fr).2 A common question is whether the chosen structure is the mostlogical one. Klaus Bollinger and I agreed when we visited thePalazetto dello Sport in 2003 that the structural system couldhave been realized in a variety of more logical ways; much of thenecessary, basic structure could have been simpler. Nervi wasobviously interested in expressing the aesthetics of structure in hisbuildings. He did not hesitate to explore the decorative, theornamental, which was a taboo for the architects at that time. Theornamentation was achieved with curved ribs, logically oriented inthe direction of the stress flow. One could imagine that thearchitects envied the engineers for this chance to deploy thedecoration for purely structurally reasons (i.e. “form followsfunction”).3 Günter Behnisch won the competition; the final design wasdeveloped in collaboration with Frei Otto.4 Model-based form-finding methods were used in earlier timestoo. Famous are Antoni Gaudí’s physical models for the church ofSagrada Familia in Barcelona. Gaudí was obsessed by finding thestructural and material given limits, which is why he investigatedevery detail in scale models.5 Harald Krejci in Frederick Kiesler, Endless House 1947–1961,Frankfurt, Vienna: Hatje Cantz Publishers, 2003, p. 12.6 At a Kiesler Symposium at the MMK (Museum of Modern Arts)in Frankfurt, Greg Lynn asserted that Kiesler did not proportionhis drawings and models. That is quite true in the sense of ageometrical and architectural proportion theory, such as theharmonic proportion theory of Andrea Palladio. But I think Kieslerdid proportion consciously his design for “The Endless House.”This difference in understanding Kiesler’s use of proportion can beillustrated with an analogous difference between the eastern ideaof music and the western harmonic theory. Whereas in westernculture composers think within a defined geometrical system ofstandardized pitches, in eastern cultures the atmosphere of thesingle sound counts, and the rhythm and time is realized as part ofthe nature. John Cage, for example, has integrated these naturalaspects of eastern music in his compositions.7 I worked as a project leader on the first three projects(Dynaform, M.art.A. Museum and Kunsthaus) in the office ofBollinger + Grohmann; the structural design for the fourth project(Klopp) was developed by osd — office for structural design (http://www.o-s-d.com) — which I founded in 2002 together with mypartners Klaus Fäth, Sigurdur Gunnarsson and Viktor Wilhelm.8 As referred to by Franken.

9 Maya is developed and marketed by Alias Wavefront (http://www.aliaswavefront.com).10 CATIA is a 3D software tool for virtual design, simulation and analysisin industrial product processes. It is developed by the French companyDassault Systems. The breakthrough of CATIA goes back to thedevelopment of the 777 airplaine by the Boeing Company in the late 1980s.This airplaine was completly designed in the CAD environment of CATIA.11 The form-generation process used in the Kunsthaus is, of all the projectsdiscussed, the most comparable to Frederick Kiesler’s. Peter Cook andColin Fournier sculpted the shape of what was later called “Friendly Alien,”as an artist would, in one piece, in difference to Gehry, who assembles hisgeometries by composition of several different components. The proportionsof the form were inspired by the site, embedded in the old part of Graz withits typical, red-colored roof landscape.12 In a digital architectural model, the geometrical data are based onsurfaces or volumes. Analytical software tools — finite elements and spatialvector framework programs — presuppose definitions of mathematical andmechanical conditions: first, an approximation of the surface or volume bya mostly triangulated pattern and, second, definition of the connectionconditions between the pattern elements. For example, points ofintersection, which are geometrically defined in an architectural model,have to be defined in analytical software with regards to their mechanicalmovabilities — degrees of freedom. Depending on the mechanicalconditions, one can more or less force a structural system into a givengeometrical form.13 “Skin deep” was articulated as such by Johan Bettum in a lecture at theStädelschule in Frankfurt.14 This is a major difference to design processes of industrial products.While high-tech products are mass-produced, nearly every building isunique, often realized under extreme pressure in a single manufacturingprocess. As a consequence, the application of innovative materials andtechniques in building construction progresses at a much slower place.However, it is only by the integration of new technologies in building thatwe are able to realize complex shapes. For example, the Experience MusicProject (EMP) in Seattle (1998–2002), designed by Frank O. Gehry andAssociates, could not have been built without the use of Global PositioningSystem (GPS) technology. The outer surface of the EMP was defined by thedigital architectural model. The distance between the outside layer of metalsheets and the waterproof layer of sprayed-on concrete underneath wasmeasured by GPS devices; the measurements provided the data forproducing the fixings.15 The description of the manufacturing process of the Dynaform project isa good example of the expense of manufacturing complex shapes; in myexperience there are, in comparison to design phases, no fundamentaldifferences in the manufacturing process between the described projects.16 This method was developed by Viktor Wilhelm, who is a partner in osd.

Page 156: B. Kolarevic-Performative Architecture

149

11

JAN EDLER

BIX AT THEKUNSTHAUS GRAZ

COMMUNICATIVE

FOR BUILDINGS:DISPLAY SKIN

Page 157: B. Kolarevic-Performative Architecture

150

11

JAN EDLER

BIX AT THEKUNSTHAUS GRAZ

COMMUNICATIVE

FOR BUILDINGS:DISPLAY SKIN

Page 158: B. Kolarevic-Performative Architecture

151

Most buildings that surround us are static — toaccommodate change, they are either reconstructed ordemolished. Generations of architects and engineershave dreamt about buildings and structures that canliterally perform in order to adapt quickly to varyingneeds or circumstances by changing the physical shape,spatial and functional configuration, levels of naturaland artificial light, overall aesthetic appearance, etc.Yet, until very recently, most of these visions havehardly been realized, primarily due to financial andtechnological limitations. Those that had been builtwere usually limited to small-scale or highly visible,high-budget projects.

Whereas the development of a physically changeable(robotic) architecture appears still to be many yearsaway, there is a promise of significant progress in digitaldisplay technologies, which allow patterns, images, text,etc., to be mapped onto a building’s surfaces, thuschanging its appearance. The decreasing costs of the “bigscreens” have already caused an increase in theapplication of electronic advertising boards on buildingfaçades. A common practice in high-density urban areasis to turn the façade into a billboard to display mostlyforeign (i.e. not building- or area-related) messages ofglobally active industrial players.

11.1The BIXcommunicativedisplay skin, designedby realities:united,integrated into themain eastern facadeof the KunsthausGraz, Austria (2003),architects Peter Cookand Colin Fournier(spacelab.uk).

Page 159: B. Kolarevic-Performative Architecture

152

SKINS THAT PERFORM:AN ARCHITECTURAL ISSUEAs the technological progress continues — and thedisplay technology becomes cheaper — the question ofwhy and how to apply this type of technology to buildingswill fall back to the realm of architecture, because thedefinition of the relation between a building and its outersurface is one of the oldest and most essential aspectsthat define architectural thinking. This will not bedifferent when the outer appearance starts to changemore rapidly (than once every twenty years). It is achange from the static “monument” to a performing“actor” that still remains in the realm of architecturalconsideration at first. “Architecture,” however, is notready for the task. Today we are missing clear andconvincing architectural concepts for the design of mediafaçades, in addition to the content that they should show.

As soon as a building façade becomes a “mediafaçade,” the ownership changes as well; this isappropriate if the façade is used for direct branding bythe company that occupies the building. But a majority ofsuch installations broadcast global advertising messages,thereby denying any form of relation between the specificbuilding and its outer appearance. The surface of thebuilding becomes separated and alienated from its inner

programmatic structure, a trend that is also evident inmore ambitious architectural or artistic experiments.

Labeled “interactive” (which is meant to stand forparticipation and democracy), the façade is turned, forinstance, into a reflective device for any passerby who,willingly or not, controls the appearance of the building,facilitated by some sensor-driven computer software.Some of that “interactivity” is appealing, but is actuallycompletely arbitrary in most cases. The currenttemperature, time, or some other banal piece ofinformation, reoccurs time after time, and is apparentlyjudged to be important enough to be broadcast at largewith considerable technical effort.

In other words, the widespread ambition to equipbuildings with media surfaces, and the available conceptsfor content, do not match. It is important to be completelyclear about the problem — it is not the financial pressureof some “evil” industry intending to conquer architecturalsurfaces to turn them into advertising billboards. Theproblem is a lack of cultural or aesthetic concepts thatare strong enough to be perceived as beneficial or evennecessary for the appearance of buildings (and, byextension, cities). We need valid dynamic aestheticconcepts — choreographies — as a continuation of thearchitectural culture that took centuries to develop!

11.2Kunsthaus Graz,as seen from theSchlossbergfortress abovethe city.

Page 160: B. Kolarevic-Performative Architecture

153

THE BIX PROJECTBIX1 is a light and media installation for theKunsthaus Graz,2 Austria (2003; figure 11.1). Ittransforms the main eastern façade of the buildinginto an alterable, performative membrane to transmitinternal processes of the art institution to the public.It is an attempt to create an experimental laboratoryfor the development and deployment of a uniqueurban communication style — a “language” — whichis synchronized with the architecture and its users onthe one hand, but which is also deferring to its urbancontext on the other.

BIX has the potential to become a referenceproject in the discourse about architecturalperformativity and the so-called media façades. It isthe first very large, permanent urban screeninstallation, run by a non-commercial entity andexclusively designed and dedicated to show artisticproductions.3 A contemporary art museum isespecially adequate as a “host” entity for this projectbecause, by its mandate, it is expected to experimentand map the potential of new communication tools,such as BIX.

BackgroundThe city of Graz in Austria was the European CulturalCapital for 2003.4 As part of that cultural project, the citycommissioned several new buildings, including the new“Kunsthaus Graz” — a museum for internationalexhibitions of modern and contemporary art, which wasinaugurated in September 2003. The building’s spectaculardesign stems from the award-winning competition designby the 1960s Archigram5 legend Peter Cook, his partnerColin Fournier, and their spacelab.uk team (figure 11.2).6

The irregularly shaped, biomorphic building structurefloats as an independent body, balloon-like, above a glassfoyer. The sleek, blue shimmering façade is the outstandingcharacteristic of this building, referred to as “FriendlyAlien.” It is constructed from more than 1,100 individuallyshaped, translucent acrylic glass panels, wrapping thewhole volume of the building like a skin (figure 11.3).

The Berlin-based architectural studio realities:unitedtook the building’s skin a step further and turned it into agiant media screen called BIX.7 The BIX concept wasinitiated and developed in the summer of 2001,8 at a timewhen the overall planning had already reached a veryadvanced stage (figure 11.4). In addition to the technicalcomplexity of the project and its advanced development

11.3Kunsthaus Graz: detailof the building skinconstructed out ofindividually shaped blueacrylic glass panels.

11.4Early (2001) conceptualrendering of the BIXfaçade (Kunsthaus Graz,2003), designersrealities:united.

Page 161: B. Kolarevic-Performative Architecture

154

11.5Exploded view ofthe BIX matrix aspart of the easternKunsthaus Grazfaçade.

11.6Installation planfor the BIXmatrix.

Page 162: B. Kolarevic-Performative Architecture

155

11.7BIX: installation of thefluorescent lamps asindependent modulesprior to the mounting ofthe building’s skin.

11.8The BIX matrixhas a lowresolution of930 pixels(lights).

11.9Resolution of theBIX matrix (whiterectangle) versus atypical TV screen(black rectangle).

11.10The size of a BIXmovie/animation —930 pixels — as ittypically appears onthe computer desktop(scale 1:1).

stage, integrating an architectural concept offoreign authorship into such an expressivebuilding design was also a challenge. After all,BIX was a new element designed to dominate thebuilding’s riverside frontage, thereby radicallyredefining the architectural concept of thebuilding’s skin. At the end of 2002, the BIXproject received approval by the client and theKunsthaus Graz architects.

DescriptionBeneath the acrylic surface facing the Mur river and thecity center, realities:united deployed a matrix of 930fluorescent light rings covering an area approximately 20m high and 40 m long (figures 11.5 and 11.6). Each lightring acts as a pixel (picture element), whose brightnesscan be computer-controlled and infinitely varied at therate of 18 frames per second. In this way, low-resolutionlight patterns can be generated over the entire façade andbe visible from a considerable distance all over the city.

Each of the matrix’s individual pixels is aconventional 40W fluorescent lamp with a diameter of 40cm (figure 11.7, and see later figures 11.13 and 11.14).The decision to use this industrial module exemplifies theasymmetrical design character of the BIX concept. Thedesign features of conventional large-screen displays wereabandoned in order to obtain a number of substantialadvantages in return.

The “resolution” of the matrix is extremely low. Thereare only 930 pixels — a mere 0.2% of the pixels found ina typical TV screen (figures 11.8–11.10). In addition, theyare monochrome only. On the one hand, such a low imageresolution imposes strong limitations; on the other,however, it enables both the modular structure and thelarge size of the installation to be highly integrated intothe architecture. The BIX installation covers virtually theentire façade facing the riverside. Using conventional “bigscreen” display technology,9 and with the same budget, thecovered surface area would have been nearly a hundredtimes smaller.

Page 163: B. Kolarevic-Performative Architecture

156

Enabling architectureSharing the same scale, the architecture and the mediainstallation together generate new aesthetic results. It isnot a separately mounted video wall but the Kunsthausitself that radiates characters and images; the projectionand the building achieve an extremely high level ofintegration as a single entity. The 930 lights of the BIXinstallation seem to be “tattooed” onto the skin of thebuilding like individual spots of pigment (figure 11.11).

The light matrix does not constitute a rectangularfield with straight sides, but rather an amorphous zonetailored to the complex shape of the building and

gradually fading away towards the edges. Hidden behind theacrylic glass façade, only the active light rings are visible,while the rest of them remain invisible behind the skin sothat the installation’s edges are not always perceptible. Thusone has the impression that the blue bubble itself rendersthe light patterns from within. In the absence of arecognizable boundary, it looks as though the light patternscould dance freely on the outside skin of the building.

BIX, however, offers the Kunsthaus Graz significantlymore than just a spectacular presentational touch becausethe installation also acts as an architectural “enabler,”10

realizing the envisioned concept of the building’s skin

11.11BIX: a new standard infusing architecture andmedia technology.

Page 164: B. Kolarevic-Performative Architecture

157

(figure 11.12). In the winning competition entry in2000, architects Cook and Fournier described the skinwith the following properties:

Much of it is opaque, but from time to time thereare revealing slivers of transparency or hints ofthe presence of action within. Strange thingsappear and disappear within the skin: signs,announcements, short sequences of film orimages: glimpsed for moments, only to fade away.For this sleek cocoon is a membrane that hints ofnew and creative activities within.11

The BIX project adopted these architectural pretensesat the time when it became clear that the initialconcept of a mostly transparent skin would have tochange for technical reasons. The installation preservesthe original design intentions, even if in a mediatedway. In this way, BIX was not only realizing the originalarchitectural vision but the installation became animportant argument in justifying the expensive acrylicfaçade panels to be mounted in front of the inner non-transparent, hermetically sealed off, bubble-like shellstructure.12

11.12By translating the originalarchitectural concept of thebuilding, BIX acts as anarchitectural “enabler.”

Page 165: B. Kolarevic-Performative Architecture

158

TechnologyThe choice to use “low tech” fluorescent light tubes(figures 11.7, 11.13 and 11.14) as the basic modulefor the display addresses the issue of technologicalsustainability over time. In comparison toarchitecture, new technologies for large screens age,i.e. become outdated, at a very fast rate. However, byusing conventional, circular fluorescent lights for theBIX pixels, known since the 1960s as kitchen lampsthat are almost a design classic today, the question ofbeing technologically up-to-date does not arise. Byusing the fluorescent light rings, i.e. an “outdated”technology, the BIX display meets the architecturaldemand of constancy. This central attribute of theinstallation — technological sustainability — savesthe operator constant upgrades and guarantees anoperational balance between architecture andtechnology at comparatively low costs.

The BIX installation is driven by custom-developed specialized software tools, which are of theutmost significance for the efficiency and precision of

the creative productions that are to be shown on thefaçade. There are two major software modules: the“BIX Director” and the “BIX Simulator.”13

The BIX Director application allows the user tocompose and schedule a program to be shown on thefaçade. The application’s interface is similar to thoseof popular video-editing environments. Four differentvideo tracks are available for arranging and mixingmultiple “events”14 on a 24-hour timeline. In thisway, the Kunsthaus Graz can set up complex “shows”for particular days or for several weeks in a row.15

The second software module — the BIXSimulator — is even more crucial for artisticproductions. It enables artists to examine the resultsin a real-time three-dimensional (3D) computersimulation of the Kunsthaus Graz in its historiccontext. By navigating through the city as if using a3D “shooter” game, artists can ensure that theirproductions adapt to the large-scale complexgeometry and the coarse resolution of the façadedisplay (figure 11.15).

11.13BIX: computerrendering showing thefluorescent light tubesas a layer beneath theacrylic glass panels.

11.14BIX: installation ofthe acrylic glasspanels over thelight tubes.

Page 166: B. Kolarevic-Performative Architecture

159

BIX AS A COMMUNICATION LABORATORYThe BIX media installation and the Kunsthaus Graz’sarchitecture share a strong symbiotic relationship.The façade as a display extends the communicationrange of the Kunsthaus Graz, complementing itsprogrammatically formulated communicativepurpose. In an abstract and mediated form, themedia façade transmits the internal processes of theKunsthaus Graz out into the public, creating asymbiosis of art, architecture and media. BIX,therefore, becomes an important identity — and animage-building factor — of the Kunsthaus Graz.

Peter Pakesch, Director of the Kunsthaus Graz,sees in BIX a new level of art mediation:

I think that the architects and especiallyrealities:united, the creators of BIX, havesucceeded in presenting a different kind oftransparency — it is not the superficial kind oftransparency of a glass house which is useless foran art museum anyway, but more a transparencyof information, the translucency of content, forwhich a lot is still to be developed and for whichthe architecture is a challenge.16

If a cultural institution like the Kunsthaus Graz is a tool forartistic articulation, the BIX installation multiplies its powerby turning the Kunsthaus Graz into a “power tool,” wherethe power is not defined in a physical sense but above all bya capacity to articulate and broadcast meaning. As thecontent producer, the Kunsthaus Graz has the chance, aswell as the responsibility, to develop methods for a dynamiccommunication between the building and its surroundings,between content and outside perception. Hence, a uniqueform of communication, consisting of vocabulary, syntaxand rhythm, needs to be developed.

At the same time, the communicating skin is a uniqueexperimental working platform for art projects investigatingforms of interaction between media and space (figure11.16). With BIX, artists can explore alternative culturaland artistic modes of production, whose implementation oncommercially used “propaganda” surfaces is widelyexcluded. The enormous size and the rough resolution of theinstallation in comparison to conventional display systemsaim at the core aspects of artistic research: reduction andintensity are well-established strategies of contemporary artto advance towards the inner essentials. In this way, BIX notonly extends the Kunsthaus Graz’s communication range —both spatially as well as temporally — but the installationreplenishes the overall program of the Kunsthaus Graz.

11.15Screenshot of theBIX Simulator3D application.

Page 167: B. Kolarevic-Performative Architecture

160

NOTES1 http://www.bix.at2 http://www.kunsthausgraz.at3 Obviously BIX is not the first non-commercial media façade.But most façade projects we know of were either temporaryinstallations or are commercial advertising boards. An overviewof interesting media façade installations and related projects canbe found on the BIX project website at http://www.bix.at4 http://www.graz03.at/5 http://www.archigram.net/6 The competition entry was designed by Peter Cook and ColinFournier with Niels Jonkhans, Mathias Osterhage, Marcos Cruzand team members Nicola Haines, Karim Hamza, AnjaLeonhäuser and Jamie Norden.7 Since 1998 realities:united has been researching intensivelythe potentials of fusing architecture and media technologies.Further information is available at http://www.realities-united.de8 In 2001, realities:united was commissioned by the building’sclient, the Kunsthaus Graz AG, to develop a “concept for thethorough integration of media technology into the Kunsthaus’architecture.” As a result, realities:united developed a broadcatalog of ideas, aiming at the creation of an overall “technicalcharacter” suitable for the functional as well as aesthetic needsof an institution such as the Kunsthaus Graz. BIX was one“small” part of this overall concept.9 LED (Light Emitting Diode) technology, etc.

10 Andreas Ruby and Ilka Ruby, “Architecture as a generalistreprogramming of reality” in ArchPlus, no. 167, 2003, Berlin: Aachen.11 spacelab.uk, London, 2000; competition entry for the KunsthausGraz.12 In 2002, the material for the outer transparent and double-curvedskin was not yet determined and different materials were being tested.At that time, an argument arose that the outer skin could beconstructed from a much less expensive non-transparent material.13 These software tools, customized for the special needs of artists,were developed in cooperation with programmers from the art scene.This approach eliminated complex translation processes during theproduction, which would have been necessary if “regular” softwarecompanies were commissioned. Software conception: Jan Edler andTim Edler (realities:united); Tobias Herre and John deKron(thisserver.de). Software programming: John deKron, Jeremy Rotsztainand Peter Castine. Available for download athttp://www.bix.at/software.14 Events are “containers” that can reference local and remote moviefiles, streaming media files, as well as specific IP addresses authorizedto “play” the BIX system remotely during a particular time slot.15 Besides this administrative function, the BIX Director is alsosuitable for artists whose work involves mixing multiple source files.16 Excerpted from an interview with Peter Pakesch in the coop99 filmproduction, Kunsthaus Graz, A Friendly Alien, 2003, Vienna, publishedby Kunsthaus Graz at Landesmuseum Joanneum GmbH, Graz.

11.16BIX as an artisticcommunicationlaboratory: videostills from the liveaudio visualperformances byartists John de Kron(Berlin) and CarstenNicolai (Berlin) forthe inauguration ofthe BIX installationin September 2003.

Page 168: B. Kolarevic-Performative Architecture

161

12

LARS SPUYBROEK

THE STRUCTUREOF VAGUENESS

Page 169: B. Kolarevic-Performative Architecture

162

12

LARS SPUYBROEK

THE STRUCTUREOF VAGUENESS

Page 170: B. Kolarevic-Performative Architecture

163

In architecture, flexibility has always been associatedwith the engagement of the building with events thatare unforeseen, with an unpredictable or at leastvariable usage of space. During modernism thatflexibility often resulted in an undeterminedarchitecture, in an averaging of program andequalization, even neutralization, of space. Ageneralized openness, as we must keep in mind,always has the effect of neutralizing events and beingunproductive because the type of space is not engagedin the emergence of events themselves. General,Miesian, openness is only suitable when all desiredevents are fully programmed in advance, by strictlyorganized bodies, as in the case of a conventioncenter, a fair or a barracks. It is flexible, open, but itis totally passive. All activity is assigned to theinstitutional body.

The architecture itself, however, does not engageinto the way events and situations emerge; it isindifferent to that. It states that life is merely theeffect of decisions that have already been takenbehind the scenes, of acts that are repetitions ofprevious acts, in which intentions are completelytransparent.

The ambition then is to find a structure, atectonics that can absorb life, chance and change,while the structure itself must last and persist overtime, to span the unforeseen with the foreseeable. Thestrategy of the Cartesian grid and the box have alwaysbeen to average out all possible events, to be generalenough for anything. Much of what we do is planned,and much of what we intend is transparent — wescript and schedule ourselves — but to engage in theunforeseen does not mean that these are just accidentshappening to our agendas. So, the problem offlexibility is not so much “to open up space to morepossibilities,” but the concept of the possible itself. Anevent is only ever categorized as possible afterwards.The possible as a category lacks any internal structurethat can relate the variations; it does not producevariation by itself — it is without potential.

The choice has always been between determinedfunctionalism and undetermined multifunctionalism,between early and late modernism, between the filled-ingrid and the not fully filled-in grid. But potential issomething else: “Potential means indeterminate yetcapable of determination … The vague always tends tobecome determinate, simply because its vagueness doesnot determine it to be vague … It is not determinatelynothing.”1 Vagueness comes before the situation;neutrality comes afterward. If it comes before, it willneutralize the forces making up the situation. We mustreplace the passive flexibility of neutrality with an activeflexibility of vagueness.

In opposition to neutrality, vagueness operateswithin a differentiated field of vectors, of tendencies,that both allow for clearly defined goals and habits foras yet undetermined actions. It allows for both formaland informal conduct. But more importantly, it alsorelates them through continuity; it puts them in a tensesituation of elasticity. This is, however, not a clean anddry coexistence of two behavioral types as a mereaddition or alternation, but more a multiplication, asone comes out of the other and shares the samecontinuum. To be able to switch between one and theother (in time) we need to materialize their in-betweenin space, clearly opposing Mies van der Rohe’s emptyopenness and replacing it with solid vagueness.

What becomes evident here is that the architectureof group behavior with all its complex dynamics isdirectly related to the architecture of the building; abehavior of continuous grouping and regrouping, ofsolidifying into certain configurations, then suddenlymelting and regrouping into other fixed states — abehavioral vagueness paralleled by an architecturalvagueness. If the skeletal structure of actions becomesas soft as cartilage and as complex as cancellous bonestructure, so does the architecture of the building. Weshould find ways where the intensive forces dealing withday-to-day decision-making and coping can actuallybecome the formative forces of the architecturalstructure.

Page 171: B. Kolarevic-Performative Architecture

164

SOFTOFFICESoftOffice2 is a building (figure 12.1) where work andplay are deeply interwoven. It is a building where bothchildren play and adults work. One half of the buildingis reserved for very young children to play withinteractive environments that are also present on theweb. The other half of the building functions as anoffice where adults work in a so-called “flexi-office”where nobody has his or her own workplace. The officeenvironment is made for both functional, formalconduct as well as for more informal creative conduct,like writing, discussions and presentations.

Intermezzo flexi-officesIn analyzing flexi-offices, it became clear to us thatthere is a real, daily tension in the effectuation of itsusage. This tension between the intended, traditional,static planning philosophies and the viable dynamicstructure is actually the force that makes it productive,not the one or the other. In calculating the requiredsurface area of an office for sixty people in verydifferent functions (marketing, administration, onlineproduction, offline production, management,origination), one would normally end up with at least1000 m2.

12.1SoftOffice,Stratford-upon-Avon, UK(2000–05),architect NOX.

Page 172: B. Kolarevic-Performative Architecture

165

Yet, within the dynamics of an office culture, if onestudies the occupancy rate of spaces that incorporatetime-space relationships, one would begin to see amuch more differentiated usage over time. Our researchgave us three categories of occupancy rates forSoftOffice: 90%, 75% and 35%. The first group wouldcontain people spending almost all of their time behindtheir own desk. The second group would contain peoplethat also spend a considerable amount of time at otherpeople’s desks and at meetings. The third group wouldcontain people traveling around considerably more thanthe others and spending much of their time in theircars, in restaurants, in hotels or at home.

This dynamic structure allowed us to make the officewith 675 m2. This means 32% of traditional planningwas excessive — a miscalculation of the efficiency dulyattributed to static thinking. But it is not just thequantifiable side of an office space that has to change;leaving the structure of the office space the same, whilereducing it a third in size, would not do any good.Leaving a standard double-loaded corridor type or officelandscape (Bürolandschaft) does not stimulate thedesired extra communication and change of behavior.Practically, then, the spaces and furniture of the officedo not need to be designated to a particular person, norstrictly designed for a particular type of work, but in

12.2SoftOffice:plan.

Page 173: B. Kolarevic-Performative Architecture

166

essence for a state of mind. We set up standard officespaces of general connectivity (flat floor, flat ceilingand general cable access) next to more informalmeeting spaces and to the very small capsules forindividual work that desires concentration (figures12.2 and 12.3). The active program is a continuum ofboth expansion (communicative behavioral types) andcontraction (the necessity to shut off, to discuss,meet, write, or shout, etc., either in small groups oralone). The passive program, more a sub-program, isone of bathrooms, cleaning rooms, editing suites andthe like.

Intermezzo: children’s spaceIn contrast with the office space, the children’s space —the SCAPE — is a space of objects; a field or landscapewhere a substantial part of the movement is propelled bymock-ups from children’s television programs (figure12.4). Where the adults in the office find a lateralfreedom in a fundamentally longitudinally oriented system,the young children’s movement in the SCAPE isgravitational and spiraling. They move “around ‘n’around” the objects. And they move from one thing toanother, from one spiral to another spiral, without anyoverview: all tension is immediately released and rebuiltagain. As most of this is brought about by the iconographyof mediated images, the architecture absorbs most of thespiraling in and out as an articulation of the floor surfaceonly. This means the architecture does not need to followthe full movement of the spiral, especially not itsrotational nature, just the fact that it is going inward oroutward, which has a slight undulating effect on both thefloor and roof surface. The rest of the children’s movementis produced by a combination of imagery and artificiallighting. Of these there are two categories: objects that arelit and objects that themselves radiate light. The lit objectsare less interactive because a pure recognition of thetelevision imagery is sufficient. For other areas a moreinteractive approach is needed: a zone where the buildingbecomes alive and starts to play with the children.3

12.3a–bSoftOffice:flexi-officespaces.

12.4SoftOffice:the SCAPE,an interactivespace forchildren.

Page 174: B. Kolarevic-Performative Architecture

167

INTENSIVE DESIGN TECHNIQUESTo map inward going and outward going forces, andto map contractive and expansive forces within onecontinuum, a networked self-organizing technique isrequired. An intensive technique means to inform avirtual system, which, during the processing of thatinformation, takes on an actual structure that is aregistering of the information. The process has totake on a highly procedural form, like cooking; theinstructions are not applied all at once, but oneafter the other, where timing becomes crucial.

An extensive top-down technique would besatisfied with cutting differently sized holes out ofsheet-like surfaces. The closed rooms that areneeded for concentration would be subtracted fromthe surface of communication. In that case it wouldhardly be possible to create continuity between bothstates. And continuity is essential if we want tensionbetween states. We are not using the expansive andcontractive as finalized properties, we read them astendencies, as working forces, as formative, not asforms. In the SoftOffice project, we again studiedthe analog computers developed by Frei Otto.

Analog computingIn the early 1990s, Frei Otto and his team at theInstitute for Lightweight Structures in Stuttgart studiedwhat they called “optimized path systems.” Previously,similar to the chain modeling technique Gaudí used forthe Sagrada Familia church (figure 12.5), they hadexperimented with material systems for calculatingform. Each of these material machines was devised sothat, through numerous interactions among its elementsover a certain timespan, the machine restructures or, asFrei Otto would say, “finds (a) form.”4 Most of themconsist of materials that process forces bytransformation, which is a special form of analogcomputing.

Since the materials function as “agents,” it isessential that they have a certain flexibility, a certainamount of freedom to act. It is also essential, however,that this freedom is limited to a certain degree set by thestructure of the machine itself. The material interactionsfrequently result in a geometry that is based on complexmaterial behavior of elasticity and variability.

In classic analog computing most of the movementis contained in gears, pistons or slots, or often in liquidsheld by rigid containers but, in the case of Frei Otto’smachines, most materials are mixtures of liquids andsolids, or they start out as liquid and end up as rigid. Inhis machines, Frei Otto often used very differentmaterials, such as sand, balloons, paper, soap film(including the famous minimal surfaces for the MunichOlympic Stadium), soap bubbles, glue, varnish, and theones that we used in the case of the SoftOffice: the wool-thread machines (figure 12.6). This last technique wasused to calculate the shape of two-dimensional citypatterns, but also of three-dimensional cancellous bonestructure or branching column systems. They are allsimilar vectorized systems that economize on the numberof paths, meaning they share a geometry of merging andbifurcating.

In the SoftOffice project, we combined a varnishtechnique and the wool-water technique (figure 12.6).The varnish technique is a surface-to-line-technique. It isbased on the effect that varnish, or lacquer, which is

12.5Form-findingthrough analogcomputing:Gaudí’s chainmodelingtechnique.

Page 175: B. Kolarevic-Performative Architecture

168

highly viscous, can also later dry up and storeinformation. For instance, one can set up a machineof lacquer surfaces that are being stretched in manydirections, break open into holes of various sizes,interconnected by threads of lacquer, which is slowlydrying up and hardening over a period of three hours.

The wool-water technique is a line-to-surfacetechnique, where the lines are given beforehand in theform of wool threads, which are set up in a patternwhere they are fixed to certain points, then given acertain amount of overlength. When the whole systemis dipped underwater and subsequently taken out,threads start to merge (which is a using-up of theoverlength), and holes next to surfaces of crossingthreads start to form.

Both techniques are fully systemic: all featuresare formed simultaneously. The holes are not takenout later but they are formed together with thevarious materializations in the system. The system iscalculating everything at the same time, solid andvoid, during the same process, through thousands ofminute iterations, where each positioning is dependenton the formation of another. Order and form areproduced, they come about, they emerge during theprocess. It is a constructivism: a soft constructivism,not a Russian mechanistic one. The constructive linesare not rigid H-beams but start as flexible rubberlines that meet up and at the end merge bottom-upinto form, into a complex inflexibility. This simplymeans we use analog computing techniques not just tocalculate structural form, but also — on a higherlevel — organizational form.

The rubber-lacquer machineIn the SoftOffice project, we started with a non-volumetric whole where all the elements wereinterconnected: a set of lines made up of rubber tubes(of 2 mm diameter) with an 8% overlength, eachattached at certain points on a rigid wooden ring (of450 mm diameter), seven points on the side of thechildren’s space and four points on the side of theoffice space. From each point there is a rubber tube

12.6SoftOffice: formfinding using acombination ofthe varnish andthe wool-watertechniques.

12.7aSoftOffice: theformation of arubber-lacqueranalog computingmachine.

12.7bSoftOffice: theoperation of arubber-lacqueranalog computingmachine.

Page 176: B. Kolarevic-Performative Architecture

169

going to each other point at the other side of thering, which makes a total of 28 lines (figure 12.7a).We doubled this system: two wooden rings each with28 tubes that not only connect one side to another,but also one ring to another. This system was thendipped into a very liquid lacquer, analogous to thewool-water technique. But while the wool-watermodel is always flat, the two wooden rings can beseparated during the hardening process. Instead ofhaving the holes and mergings in a flat configuration,we can now calculate curvature of the rubber tubestogether with the intermediary curvature of thedrying lacquer in a spatial configuration. Theseparation of the wooden rings during a three-hourprocedure is analogous to the splitting of the floorand ceiling (figure 12.7b).

So, while calculating programmatic forces —mental states — we are also calculating structuralforces. There is complete vagueness: never fullycolumn, never fully wall, never fully floor (figures12.3 and 12.4). The system is negotiating everythingwith everything without resorting to equalization.Now, what becomes most prominent in this system isthat there is both an expression of rigidity and offlexibility. We found a methodology that allows us tocalculate the in-between, to be able to vary betweenthe two states, regardless of how opposite they are.While all flexibility is expressed in the middle of thesystem, the rigidity is produced close to the woodenrings, at the edge of the system. Type is at the edges,diagram in the middle; full bottom-up in the middle,full top-down at the edges — spatially: a spongyporous structure in the middle zone, clean separationof floor and ceiling at the edges (with columns in-between). In the SCAPE, this clean separation meansthe hall-like tendency of the structure; at the officeside, four separate ‘fingers’ with gardens. It is nowthe in-between that becomes operative: it is not just aCartesian choice, it is an actual sense of tension, amaterial state of in-between that is internalized, thatbecomes effectuated in daily behavior andfunctioning.

Soft rigidityThe first step in the previously described process containsonly geometry, no materiality; the materiality then takesover during a stage of reshifting and the procedure comesto a halt in a state of full geometry again, but a geometrythat is now not imposed on a material but is the result ofmaterial interactions. It starts out explicitly Euclidean,but it does not finish as such because at the end there is noclear segmentation of dimensions any more. While wecould call the first step of the system a geometricalsurface, a system where all directions are equally present,the final stage of the model is much more complex becauseit consists of patches of crossings, mergings and holes. Thecrossing patches consist of two dimensions, which meansthat, in these areas, many directions are still available inthe system — many lines keep on crisscrossing each other,similar to the initial state. The merging patches consistonly of one dimension, where the system takes on onesingle direction — many lines stick together to form amain artery. And the holes, of course, are areas where welose all dimensions and no directions are availableanymore.

While the first stage consists of homogeneous tiling,the last stage consists of heterogeneously-nested patching.The end result (figure 12.7b) is based on looseness, but isitself not loose and not weak, but rigid and completelytight (when attached on an open ring it comes out of thewater straight and horizontal). It is a strategy of flexible,individually weak elements cooperating to form strongcollective configurations.

What emerges is a complex or soft rigidity, which isvery different from the top-down, simple and frozenrigidity of the first stage. We should therefore resist theidea that the first stage is a rigid order and the end resultis just a romantic labyrinth or park. Actually, thearabesque order of the end result is as rigid as the firststage of the grid, but much more intelligent because itoptimizes between individual necessities and collectiveeconomy. Yet it is not an easily readable and clear form oforder, but a vague order; it is hardly possible todistinguish between surface areas, linear elements andholes. Surfaces can function as linearities, lines can

Page 177: B. Kolarevic-Performative Architecture

170

cooperate in surfaces, and holes can exist at allscales. Everything between the dimensions ismaterialized. Although the dimensions are clearlysingularities arranging the system (the mergingsinto thick lines are like the ridges of dunes, whichorient the sand surface to the wind forces), it iscontinuity that makes them emerge — although theorder is vague, it should nonetheless be consideredvery precise because nothing is left out. There is norandomness; there is only variation.

The interesting feature of this system is that it isin fact structured by holes; the nesting of holes isthe driving force behind its formation, whilearchitects are always trained to think that holes are,in the end, subtracted from a system. This machinedoes not operate on subtraction or addition, but onmultiplication, in the classic sense of early systemstheory, which states that a whole is always largerthan the sum of its parts. Here porosity is anemergent property. The first stage (figure 12.7a) isbasically drawn, contrary to the end stage (figure12.7b), which is processed by a machine, i.e.calculated. All effects that coexist in the final result,all the curves, all the mergings, all the holes areinterrelated; nothing can be changed withoutaffecting the arrangement of the whole. All lines aremobilized simultaneously, in parallel, while drawingis serial; one line is drawn after the other. A drawingis always created in the visual field, while the analogmachine follows a partly blind and informationallogic where the image is the end product of theprocess. And although this technique should beconsidered as a hybrid of the top-down and thebottom-up, the drawn and the generated, itsintelligence lies in the fact that nothing is“translated;” the drawn is not “translated” into thereal. In itself it works in one-to-one scale — in thatsense it is not even a model. This direct proportion isone of the main features of analog computing, whichdoes not simulate by numbers but by an empiricalrescaling of the real.

The organizational and informational stage is material, notimmaterial, as is so often put forth. It is the materialpotential, the material distributed intelligence which sets themachine in motion — a transfer of water-turbulence to wool-curvature. Then it is the stickiness, the hairiness and thecurvability of the wool thread together with the cohesiveforces on the water surface that bring it to a halt again andinform the end result. It is an intensive technique within anextensive system, and though the quantities (surface area,etc.) are given beforehand, the quality emerges through theinteraction and multiplication of different parameters.Generally, the intensive is a deformational property (likeheating), but here it also becomes a transformationalproperty (like boiling): the threads restructure andreorganize to “find form.” The system as a whole passes acritical threshold. The degrees of freedom of deformation,which are more like extensive movements within an internalstructure, become intensive, qualitative changes of “that”structure.

Wet grid versus dry gridThe classic regular, Greek grid is a system that separatesinfrastructural movement from material structure. Simplyput, the structure is of a solid, while the movement is of aliquid. We must consider the orthogonal grid as a frozencondition because its geometrical state of homogeneityrelates directly to a material, crystallized state offrozenness. Frozen states are simple states and, of course,these have been the first to be mastered by the geometers,but to understand complex states we need to developcomplex geometries. Generally, we are taught to think thatgeometry is the higher, the more abstract and pure form ofmateriality, which is a misconception because althoughgeometry urges for the necessary exactitude, it is totallyimprecise. Any geometer comes after the event, wheneverything has dried up and, therefore, he can only bedealing with the extensive state of the material, taking uplength, width and height.

The wet grid, Frei Otto’s grid, is one in which movementis structurally absorbed by the system; it is a combination ofintensive and extensive movement, of flexibility and motion.The geometry does not follow the event, geometry co-evolves

Page 178: B. Kolarevic-Performative Architecture

171

with materiality; it is generated through analog, wetcomputing. One could call the organization of the finalstage wet and its structure dry. While it itself is notmoving anymore, it has attained an architecture ofmovement. In this sense, movement must be viewed asinformation, as pure difference, because we all knowthat when “information” does not cause any change itis superfluous. It simply did not “in-form,” it did notenter the form. This means movement in itself is notenough to be called information, it must be internallyprocessed as a (temporary or permanent)transformation. The physical displacement ofmovement must be processed as a structural change.

Basically, emphasis on movement as deformationis merely indexical and is meaningless when notresulting in structural transformation. The freezing ofmovement becomes merely traces, momentarystoppages of a bygone present; they are not, however,structured through time, they are not paths whichallow for movement to be rerun over and over again,and slowly condense and evolve. On the other hand,they are not roads either, which with their exactdistinction between surface and line prevent thesystem from reconfiguration and adaptation. Eachstate of path-forming should function as the analogcomputer of the next one. There should be enoughsolidifying for registering and there should be enoughplasticity to enable changes. This brings the optimizedpath systems of Frei Otto close to contemporary multi-agent computing devices based on ant colonies withtheir pheromone distribution.

For a real-time, analog computing model we needtwo things: first a system that is internally structured(or else it cannot process information) and, second,external flows of information. This simply means thereare always double states, simple states and complexstates, coexisting in gradation. Higher states ofinformation can only happen in lower states ofinformation, they coexist hierarchically but within acontinuum. They do not exist next to each other — thegeneric and the specific share the same continuous,topological space as do the standard and the non-

standard. One is always engulfing the other: we need tostart from a state of equilibrium that already containsinformation through its structure, create disequilibriumto increase the amount of information, and then we needequilibrium again to memorize it.

The brilliance of the Frei Otto’s model is that theflexibility is taken literally and materially, that the realmovement of water flow becomes the abstract movementof wool-structure, which results in a coherent language of“bending,” “splitting,” “curving,” “nesting,”“aligning,” “merging” and the like. All arabesquefigures in the final state of the model immediately relateto complex configurations.

obliqueWTC5

Undeniably, the skyscraper is the most successful buildingtype of the twentieth century. Its generic reductionism,however, its passive stacking of human behavior, itsmanic monoprogramming will and should becomeobsolete and, as a type, it will have to be rethought,making a new evolutionary step of the megabuildingpossible. We should try to find urban strategies to dealwith the “Huge,” with global forces working on localsituations. We should find ways to work against thehomogeneous and find other ways that are more open tolife, the changes of life, and the unpredictability of life.

In rethinking the mega, the Huge (which is slightlyrelated to the “High”), we should be more concernedwith the structure of the Huge than its size. Developingtechniques to heterogenize the Huge, without just simplychopping it up or collaging it together, means we have toreconsider the design techniques. Moving away from top-down towards bottom-up techniques — especially whendealing with superlarge structures — becomes absolutelynecessary.

In developing our design for the rebuilding of theWorld Trade Center (WTC) in New York, we reused theold wool thread modeling technique invented by Frei Ottoand described in previous sections. We used one woolthread for each core of the destroyed or damagedbuildings on the former WTC site. In an inverted model,the wool threads hang straight down under the sole

Page 179: B. Kolarevic-Performative Architecture

172

influence of gravity forces. When dipped into water andtaken out again, all threads reorganize themselves into acomplex network6 (figure 12.8), comparable to bonestructure. The structure is no longer formed by a simpleextrusion of a plan but instead self-organizes into anetworked megastructure, where the whole is larger thanthe sum of its parts.

We thicken each of the wool threads into a lean towerthat merges and splits up as it moves upwards (figure12.9). This enables the structure to comply with the NewYork zoning law that only allows high buildings to occupy25% of the total site surface area. In this case, however,the 25% is always positioned somewhere else, making itboth into one single megabuilding with many (structural)holes in it or many thin towers that cooperate into onelarge structure. The towers sometimes act as a bridge,sometimes as a counter-structure for another one, andsometimes free themselves to become a smaller sub-tower.Most of the loads are carried through the honeycomb steelstructure of the surface, helped by an interior column grid,which follows the diagonals of the towers similar to LouisKahn’s design for the Philadelphia City Hall from 1957.Also, the elevators form a highly complex structure ofdiagonals where, at some platforms, more than five or sixdifferent cores come together to form larger public areas.It is this network of elevators which makes the buildingnot just a new type of tower but more like a new type ofurbanism. The elevators become an urban extension to thesubway system: a punctuation of the street by atechnological system to intensify its public functioning.Generally, all interactions of a Manhattan block (with itsprogrammatic diversity that should at least be rivaled bythis new building) only happen on the street, while allbuildings blindly tower away from that level into a non-interactive side-by-sidedness. Here we re-network thestreet into the tower. We read the wool-thread diagramboth structurally and programmatically, where thestructural ‘diagonals’ become a reemerging of Virilio’soblique:7 lateral, horizontal street forces are multipliedwith the vertical stacking model of the skyscraper,resulting in an oblique tower.8

12.8Form-finding:World TradeCenter, New York(for Max ProtetchGallery, 2001),architect NOX.

Page 180: B. Kolarevic-Performative Architecture

173

12.9World TradeCenter, NewYork (for MaxProtetch Gallery,2001), architectNOX.

Page 181: B. Kolarevic-Performative Architecture

174

A SOFT CONSTRUCTIVISMThe techniques invented and suggested by Frei Ottohave been very diverse, varying between theapplication of already invented techniques to ongoingprojects and more fundamental research into materialform-finding. Not surprisingly, his optimized pathsystem machine is quite unique within the whole of hisresearch because he rarely bothered with horizontalstructures. Essentially, his research was in thecomplexity of the elevation, the structure and not theplan.9

Patterning effects and configurational, emergenteffects happen on all stages, both in the plan and inthe elevation. Instead of following the plan–floor/extrusion–wall method, we should opt for a methodwhere elevation and plan become more intertwinedand coevolve into structure. For centuries, the orderwithin the design process has been: first the plan(action), then structure at the corners (construction),which at the end is filled in with walls (perception),where the latter two have been part of the splendidSemperian distinction between tectonics and textile.10

Our agenda should be to short-circuit action,perception and construction. Having weak textilethreads team up into rigid collective configurations isa direct upgrade or inversion of the Semperianparadigm. But they should be three-dimensional fromthe start; plan threads can twist and become wallthreads.

All these techniques already exist in textile art,where complex interlacings occur in crochet, weavingand knitting. The art of the arabesque is as old asarchitecture; it has just never been conceived at thescale of the building. And that is certainly because oftechnological reasons — the arabesque has alwaysbeen accommodated by manual labor while thestraight extrusion was necessarily associated withstandardization and industrialism. Clearly that ischanging with non-standard architecture.

We should be careful though not to mistake the non-standard for “free-form architecture,” for theamorphous or even the streamlined; we should strive fora rigorous non-standardization, rethinking repetitionwithin sets of variability, rethinking structures withinranges of flexibility. The more we move towards thenon-standard, the more articulation must become anissue. If there is no technology of design, a technologyof manufacturing becomes nonsensical. With machinesunder numerical control, we need the design processitself as an informational procedure also, with clearlystated rules and scripts to generate a structure ofvagueness.

We have argued here and before that starting withthe soft and ending with the rigid will offer us muchmore complexity in architecture. And here we are notreferring to Venturi’s linguistic complexity (ofambiguity)11 but to a material complexity (ofvagueness). Obviously, the science of complexity hasproduced many diagrams of the soft, and these haveoften been dropped onto rigid architectural structuresor typologies. Although deconstructivism proved to besuccessful in breaking down most of the top-downordering tools we were used to in architecture (contour-tracing, proportion, axiality, etc.), it proved to beincapable of instrumentalizing complexity itself as atool that was material and architectural. It understoodevery act of building as an implicit counter-act, as anegation — meanwhile, the engineers silently repairedit.

We should understand all objects as being part of aprocess of emergence; the made as being part of themaking, not the unmade. Our goal must beconstructivism, or emergence, and anything thatemerges should coemerge; the way we see is emergent,the way we move around, the way we act in relation toothers, to our habits, to our memories, all theseemergent patterns should coemerge with its materialstructure.

Page 182: B. Kolarevic-Performative Architecture

175

This makes our agenda one of a post-industrialconstructivism, a non-standard constructivism. Allbehavior is material, all structure is material. “Howdo we orient, how do we feel, how do we group orungroup?” all these questions should be posedsimultaneously, together with “how does it standup?” There have been many attempts to borrow“images of complexity” that were fed into eithercirculational, formal or structural diagrams; Kleinbottles, weather maps and the like are interestingbut are not enough. We should create complexity byfeeding them into each other. We should feedcirculation into structure, feed structure intoperception, and feed perception into circulation. Itdoes not matter where we start as long as we arelooping a flexibility of action (affordances) into aflexibility of structure (vagueness) into a flexibilityof perception (atmosphere), looping non-standardbehavior into non-standard structure into non-standard architecture.

Page 183: B. Kolarevic-Performative Architecture

176

NOTES1 C. S. Peirce (ed. Peirce Edition Project et al.), The EssentialPeirce: Selected Philosophical Writings, vol. 2, Bloomington,IN: University of Indiana Press, 1998, pp. 323–324.2 Shop, interactive playground and headquarters office forRagdoll TV Productions in Stratford-upon-Avon, UK (2000–05),architects NOX (Lars Spuybroek with Chris Seung-Woo Yoo,Kris Mun, Florent Rougemont and Ludovica Tramontin).3 The area is what we have called “Glob.” Glob is a worlddesigned by globally networked children. Glob is both present inthe building and on a website. Glob is a “living organism” (someof its responses are calculated with genetic algorithms) that hasthe special ability to interact with children. Glob will grow, loveand, of course, play. Glob creates an experiential environment forthe children that touches upon their senses and humor. Togetherthey will create drawings, music, stories and love.4 Frei Otto refers to this on many occasions as Form Findung —finding form or taking form. Together with Bodo Rasch, hepublished a book with the same title (Finding Form, Munich:Edition Axel Menges, 1995). The terminology in German wasnot used before Frei Otto coined it. One should frame theformgebung-formfindung debate in a German discussion goingback far earlier, especially the one during the late eighteenthcentury on Preformation and Epigenesis. For a more elaboratediscussion see Detlef Mertins, “Bioconstructivisms” in LarsSpuybroek, Machining Architecture, London: Thames andHudson, 2004 (forthcoming).5 NOX (Lars Spuybroek with Chris Seung-woo Yoo and KrisMun) for the Max Protetch Gallery, New York, 2001.6 The cohesive lateral forces of the water are added to thegravitational system.

7 Paul Virilio and Claude Parent, Architecture Principe, Paris: Leséditions de l’imprimeur, 1997.8 We included a Memorial Hall inside the building. High up in thestructure several floors are taken out to form a large open spacethat gives an open view over the city, but the space will also bevisible from most areas around New York. The hall should not be amonumental petrification of mourning but should be a projectionspace where visitors can interactively request for home videos,photographs and websites of the all people lost, and meet them.9 Frei Otto was always invited to cooperate with architects that hadalready developed the plan, and his contribution was subsequently inthe typical engineering stages, afterward.10 In Die Vier Elemente der Baukunst (The Four Elements ofArchitecture, London, 1851), Gottfried Semper made his famouscategorization of the elements of architecture through a materialclassification: a. earthwork/foundation; b. tectonics/wooden poles atthe corners; c. textile/wall as an infill of the wooden poles; and d.the hearth. In the same vein I always add: floor/action, corner/construction, wall/perception, and fire/sensation. Semper’scategorization is based on his seeing of the Caribbean Hut duringthe Great Exhibition in London where the structure consisted ofload-bearing columns of wood on the corners, with walls made ofthe non-load-bearing element of woven textile. Although we herefully underline the implicit notion of surface and textile as the mainconstituent element of architecture, we can basically considertextile now as a tectonic element in its own right, especially whilereferencing to Frei Otto, not so much to his application of textile intent architecture but through his methodological use of textile in theanalog computing of form.11 Robert Venturi, Complexity and Contradiction in Architecture,New York: Museum of Modern Art, 1977.

Page 184: B. Kolarevic-Performative Architecture

177

13

ALI RAHIM

OPTIMIZATION INARCHITECTURE

PERFORMATIVITY:

EFFICIENCY ANDBEYOND

Page 185: B. Kolarevic-Performative Architecture

178

13

ALI RAHIM

OPTIMIZATION INARCHITECTURE

PERFORMATIVITY:

EFFICIENCY ANDBEYOND

Page 186: B. Kolarevic-Performative Architecture

179

Technical articles of research tend to use the term“performance” but rarely define its meaning. Howdoes the performance of a computer relate to theperformance of a material system? There is nogeneral definition of performance.1 The precision ofeach description comes from its local context. Theperformance of a technology, for instance, refers toits technical effectiveness in a specific evaluation orin a set of applications.

Technical disciplines are constituted arounddevices conceived as essentially functional andtherefore inevitably oriented towards efficiency. Thisobsession with efficiency is prevalent throughoutsociety and is reflected in the design of many devicesand systems, but it has not demonstrated anunderstanding of technology as it has existedhistorically, as it exists today, and as it may exist inthe future. Technology is not merely technical; it is anactive and transformative entity resulting in new anddifferent cultural effects.2 Technology in this sense isnot an efficiency-oriented practice, measured byquantities, but a qualitative set of relationships thatinteract with cultural stimuli, resulting in behaviors,some of which are techniques.

Our work uses techniques of design that areprocess-driven and that produce these transformativeeffects in cultural, organizational, technological,social and political production. Techniques havealways contributed to the production of human andcultural artifacts, but their refinement andacceleration after the industrial revolution hasemerged as the single most important element in theevolution of cultural endeavors.3 Techniques influencethe design of objects, which, in turn, produce effectsthat influence human behaviors and produce newtechniques that affect an object’s technicalperformance. This results in the selection of newpossible objects which generate further cycles ofdevelopment.

PERFORMATIVITYThe path of evolution produced by a cultural entity — anobject, a building, a company or a career immersed in itscontext — produces a distinct lineage4 as a result of itspropagation. Each lineage exists indefinitely through time,and may be selected in terms of its performance. Althoughdifferent performance paradigms are separate, theyinevitably emanate and influence each other. This selectionalways constitutes an emergent and decentralized process inwhich many actors are interrelated through complexfeedback loops. This process requires cognition, isinnovative, and leads to a restless proliferation of neweffects. This transmission across lineages is, as Stephen J.Gould writes, “the major source of cultural change.”5

Animation techniques enable us to inhabit thesepotentials in time and cross these lineages. Our designprocess reacts to external stimuli and transforms a situationthrough feedback between a subject and the environmentand between architecture and its milieu. The material,organizational and cultural change that occurs as a resultof this perpetual feedback and two-way transfer ofinformation is performativity. Here, models developed inone research paradigm can be appropriated by another.These paradigmatic readings not only have the ability togenerate, describe and evaluate performances, they also citeand recite them, breaking apart the evaluative forces thatbind together their discourses and practices. Moreimportantly, they can recombine and reinscribe these forms,deploying them elsewhere while incorporating, ignoring orreevaluating their values in other ways.

Performativity always has the potential to produce aneffect at any moment in time. The mechanisms ofperformativity are nomadic and flexible instead ofsedentary and rigid. Its spaces are networked and digitalrather than enclosed, and its temporalities are polyrhythmicand non-linear. Performativity produces new subjects ofknowledge, hyphenated identities, transgendered bodies anddigital avatars. The performative subject is fragmentedrather than unified, decentralized rather than centered,virtual as well as actual.

Page 187: B. Kolarevic-Performative Architecture

180

In our winning competition entry for a Leisure Generator,sponsored by the Greek Government in Association with theHellenic Ministry of Culture for the 2004 Olympic Games,our design strategy is to sample, providing the greatestrange of potential effects that different cultures mightproduce in interacting with a surface and one another. Thedesign celebrates the heterogeneity of all world cultures byintensifying different leisure activities and cross-pollinatingsporting, social, recreational and domestic leisurecategories during different periods of the day (figure 13.1).

We allow for many leisure spaces to occursimultaneously in order to stimulate unforeseenrelationships that might emerge through the feedbackbetween bodies and the inhabitable surface of the project.Here, different cultures are homogenized by the body ratherthan by color, gender or religion. Some bodies may sit on asurface while others may use the same surface for laying orreading. Individual users may influence the person next tothem, affecting the way that groups form and interact withthe surface. These relationships become more direct at themore particular geometries; for example, stairs can be usedfor walking, sitting or sunbathing (figure 13.2).

The structure is redundant and built in an accumulativemanner. It eliminates thresholds and moves them towards agradient where one can have different levels of opacity inone surface or different lighting qualities within a singlespace. Opaque, translucent and transparent effects canoccur in one surface in continuous variation. The conflationof structure and material combine to change relationshipsand recombine to produce effects that are guided by theformation. Here, the detail is everywhere. This non-reducible geometry allows the subject to move through thisfield with a degree of ambiguity, avoiding conventionalnorms of expression and prescribed interpretation.

Performativity influences the outcome of habitational,material and ambient effects perceived by users and theeffects they have on their milieu by reconstituting oursensibility of architecture (figure 13.3). These formationshave no bounded limits in figure or ground, building orlandscape, inside or outside, public or private, but provide acontinuous gradient between extremes and a rich palette ofpossible interactions. This constitutes an unfolding field of

13.1Plan view showingdifferent inflections acrossthe inhabitable surface ofthe Leisure Generator(2003), architect AliRahim/CAP.

Page 188: B. Kolarevic-Performative Architecture

181

social discourses facilitated by geometric articulation. Ourintention is to catalyze interactions between the formationand the subject by opening the work up to a wide field ofpossible orders rather than a single definitive one.

The emergence of the computer is another example ofthis potential. It is clear that each technological lineageembodied in the computer encompasses the contributionsof scholars, philosophers, visionaries, inventors, engineers,mathematicians, physicists, and technicians. Each lineageis stimulated by vision, need, experience, competence, andcompetition. As these lineages develop through time, theyorganize effects already in existence, such as the use ofmachines and automation. They organize advances madein symbolic logic and in science and mathematics, whichbecome feasible with the invention of numerical patterningsystems. These factors are impacted by different intensitiesof economic, commercial, scientific, political and militarypressures, crossing the technical threshold and emerginginto the temporally organized technological formation thatis a computer. The computer here is an effect ofhybridization across these lineages.

If we were to view these non-linear organizationalprocesses as fixed in space and time, the resulting objectswould be severely limited and would strain to representmeaning through formal expression. This object type ispassive and defined only by its material attributes whichare linear and causal. Such an object is static and only hasthe capacity to produce predetermined effects. To avoidthis stasis, we must view the object in its context andunderstand it as a part of a continuous temporal process.For example, several types of machines, including

13.2Leisure Generator:detailed plan view showingthe accumulation of thestructure through time.

13.3Plan of landscape andbuilding at Variations,London (2002), architectAli Rahim/CAP. The formof the building andlandscape are developedsimultaneously as one isan intensification of theother.

Page 189: B. Kolarevic-Performative Architecture

182

tabulators, sorters, transcribers, and verifiers, werecross-affiliated to produce the computer.6 If eachmachine is viewed as being fixed in space and time, thenthe sorter, for example, would become a better sortingmachine and not part of the organizational process thatinfluenced the emergence of the computer. This cross-affiliated process operates from within, is influenced byits environment and has a potential to spontaneously self-assemble and produce effects which are qualitativelylarger than initially anticipated.

The computer is instrumental in affecting its users.They are influenced by their interactivity with it and, inturn, impact their environment. This feeds into a cycle ofproductive emergent effects which influence our culturalmilieu. This process is temporal and each point in timebecomes a potential generative juncture through whichtemporality is established.

A desire to study the behavior of matter in its fullcomplexity brings with it an ordering of the world wherematerial systems are understood to be similarly cross-affiliated and in a constant state of flux. The analyticalapproach to design, which studies a site or project onlyfrom the top down, misses the properties that emergefrom complex interactions between parts, as in themodernist notion of the diagram as building parti.

Our work seeks to harness the potentials ofperformativity by using temporal techniques to simulatethe indeterminable unforeseeable influences andinevitable unforeseen events and accidents that invigoratedynamic environments. We study the behaviors ofanimations over time, analyzing both quantitativechanges and qualitative rates of change. We guide andshape formations from within this process, using thecomputer’s ability to generate and regeneratepossibilities iteratively, adjusting and fine-tuningrelationships between matter and use.

The resulting formation is itself new, with newproperties that do not belong to the old. Just as water isnot a property of hydrogen and oxygen, and a tornado isnot a version of the winds that precede it,7 inorganicmatter is much more variable and creative than we everimagined. One way to fully participate within complexenvironments is to develop a spatial-temporal formationthat negotiates the real-time response of the system thathelps generate it. Here the traditional axis of hypothesis,analysis and intervention gives way to perpetual feedbackbetween analysis, intervention and exchange with theenvironment time and time again (figure 13.4).

13.4Variations: entry— there is nothreshold betweenbuilding andlandscape, but acontinuousgradient in-between.

13.5Variations: anexterior view ofthe buildingshowing itsformal variation.

Page 190: B. Kolarevic-Performative Architecture

183

This dynamic exchange between intervention andenvironment is studied in Variations, a weekendresidence for a fashion designer in London (2002).The project is also a venue for fashion shows heldseasonally for her best clients. Here, the notion of aquantifiably efficient design looses its meaning(figure 13.5). There is no optimal fixed solutiononce and for all, as these criteria change over time.Habitational potential emerges from the spatial andmaterial attributes that constitute the project’sform. No space is described as a function of its useor by event, but of the qualities it contains.

Animation techniques were used to simulateperformative effects (dining, sleep, spectatorship,work, etc.) at different scales through an iterativeprocess (figure 13.6). One scale examined thefeedback between building and environment

13.6Variations: theinflections in thesurfaces influenceto different extentshow the individualuses each space,and how this affectsgroup use.

Page 191: B. Kolarevic-Performative Architecture

184

(drainage, vegetation), the second between the eventsschedule and the form of the building and landscape,and a third scale studied shifting scenarios that emergeon a daily, weekly and seasonal basis. The actualhabitation of the space emerges through negotiationswith the surface by users and may vary with differentsocial situations (figure 13.7).

Several tendencies emerge through thisinteraction. These are both specific and charged incertain places and primarily unaltered in others.Patterns of use include crowding or dispersion and

always shift between extremes in the same space.Redundancy within the structural strategy of the projectprovides for greater formal variation, as does redundancyin the surface articulation. For example, in the openings,we studied the patterning of light through different timesof the day and year (figure 13.8). We wanted to intensifythe gradient between different conditions, such as lightand dark, smooth and rough, loud and quiet. We alsoprovided for multiple routes through the project, somefast, others slow, with the intention of giving the formgreater potential to be used differently.

13.7Variations: threespeeds of circulationare brought togetherthrough an inflectedsurface of differentextents.

Page 192: B. Kolarevic-Performative Architecture

185

13.8Variations:the lightinggradient.

Page 193: B. Kolarevic-Performative Architecture

186

These potentials are not achieved through efficiency.Instead, they are products of performativity wheredesign, manufacturing and optimization are seen not asstatic elements within the development of a buildingbut as participants in a dynamic model and as part of atemporal process. This shifts optimization andmanufacturing from being efficiency-based practices,concentrated on quantities and the production of staticobjects, towards being conditioned through theirdevelopment and evolving possibilities for new spatial-temporal qualities and experiences. These effects areseamless between environment, user, formation,structure and material distribution.

BEYOND EFFICIENCY AND OPTIMIZATIONOptimization often points towards the most efficient orstructurally minimal solutions. While non-linearanalytical software is available, it is typically used on aproject post-design to make sure a structural andmechanical system is viable. This model of optimizationalways seeks to maximize the efficiency of a form,searching for the purest solution in an attempt to domore with less material, move air better, etc. Howmuch more efficient can a design get? How muchthinner can a cable get? We can already imagine thatdue to advances in material science, in fifty years thesmallest tensile member might be as thin as a hair, butthe reduction of material quantity to its minimum, orthe reduction of a form to its most efficient structuralexpression, would fail to realize the relational andformal potential of the new materials. By all evidence,these techniques of efficiency and reduction from thetop down have reached their limit and, in all likelihood,they will be unable to produce new possibilities forinhabitation or to generate new effects.

The software being developed in advancedengineering offices is able to generate and test formsdynamically. This software is pointed in the correctdirection but its focus needs to shift from structuraland mechanical optimization towards a way of evolvingnew possibilities for spatial-temporal experiences andmaterial potential in design and manufacturing. The

manufacturing industry needs to become more flexible andadaptive, and shift towards a model of performativity. Thiswould entail evolving a building through performativity andsimultaneously testing its effects in different social settings.This would allow designers to develop and test theirformations immediately in the pursuit of producing effectsand environments that have habitational potential. It wouldalso forecast unpredictable conditions and would inform usin real-time about our formal, material and ambient effects.

Incorporating manufacturing within this model wouldshift production processes from mass production and masscustomization to one where the dynamic evolution andproduction of non-prescribed possibilities is possible. Somemanufacturers, like Nike in its NikeID line, are shiftingtowards personalization as a manufacturing process, butthey still rely on postmodern collage techniques bycontrolling the number of possible outcomes. They simplyprovide a wider matrix through variety that customers canselect from instead of a more fluid variability.

We can look towards the automotive industry forinspiration and a possible direction for developingtechniques based on greater temporal redundancy. Here,there are many scales of redundancy. For example, crashtesting of the automobile chassis is done both virtually andphysically, and in as many scenarios as possible. Automobiledesigners employ an iterative process and come acrosseffects that they can predict and others they cannot foresee.These effects are generated by the variables that producethem and it is this outcome of effect that highlights thesevirtual techniques and reveals the potential for them to beapplied in architecture and other design disciplines. Thispotential arises from the fact that the total environment isincorporated within the software in order to produce adynamic testing ground for material and effect. Virtualcrash testing is very specific. Speed, material properties andenvironmental conditions are all coded within the software,which uses magnitude of force and direction to anticipatedeformations in the vehicle and injuries to the passengers.This allows for highly precise calibrations within an evolvingperformative feedback structure that has the capacity togenerate and test emergent effects on users throughout thedesign and manufacturing processes.

Page 194: B. Kolarevic-Performative Architecture

187

Our Catalytic Furnishings project (2003) pointstowards the potential of a flexible manufacturingtechnique that evolves through the design process.We were interested in the way the body acts in waysthat we anticipate as well as ways it might actdifferently depending on interactions with the surfaceof a furnishings element. We aimed to charge asimple material surface with varied potential foroccupation and use.

The project mixes and samples several different lifestyletypes and uses them to understand primary bodyperformances on different orientations and surface qualities(figure 13.9), from rigid vertical surfaces for leaning tosofter more horizontal surfaces for laying, and working tocompressible horizontal surfaces for sunbathing. Theheterogeneous geometry of the pieces tends to destabilizesimple division between individually occupiable spaces andnecessitates an ongoing territorial negotiation between

13.9CatalyticFurnishings:lifestylediagram.

Page 195: B. Kolarevic-Performative Architecture

188

simultaneous users. Users input variables of their lifestyleinto software that deforms and shifts a structural cage(figure 13.10) and the manufacturer has the ability todynamically vary the form of a mold in real time (figure13.11). This variable mold system we are developing,called FlexiMold©, allows for cross-affiliations ofdifferent body performances to be registered. Through aperformative process of selection, users are able to createtheir own customized object that responds to theirparticular lifestyle.

13.10Catalyticfurnishings: cagedeformations —users inputvariables of theirlifestyle intosoftware thatdeforms and shiftsa structural cage.

13.11Catalyticfurnishings:performancevariationdiagram.

Page 196: B. Kolarevic-Performative Architecture

189

We realized that if through cross-affiliation no singlebody position was integrated, we would have thegreatest success. The body would be inflected inpositions it was not accustomed to due to interactivitywith the surface. Multiple users on the same surfacemight produce interesting or even compromisinginteractions between themselves, and the furnishingwould engender different social arrangementsdepending on its location (figure 13.12).

Through the design process, we limited the number ofvariations (figure 13.13). This informed the design of theFlexiNurb machine and the spline structure. The formsthat are produced are not referential to existing furnituretypes, but they might be considered to be loosely related tothem. Each variation is possible within the streamlinedmanufacturing process we have devised. A standardthickness of material, which has been designed for theworst-case structural scenario, is placed in the mold. The

13.12CatalyticFurnishings:scenarios ofdeployment.

Page 197: B. Kolarevic-Performative Architecture

190

material is redistributed in the mold, producing avariable cross-sectional thickness. The thickness is notdetermined by structural efficiency, but by the effectsthat are desired (figure 13.14).

All possible variations are structured usingfiberglass monocoque shells and are available in a widerange of finishes (figures 13.15 and 13.16). In onevariation, the fiberglass thickens to allow for a gelinsert, which changes the quality of the surface. Gels ofvarying thickness provide further gradients of softness.In another variation, we use pressure-sensitive foamthat responds to body weight and allows for a differenttype of variation. Performative iteration is thusoperative at a variety of scales, inflecting thenegotiation between user and surface, and cross-profiling these inflections to different extents and inways that produce the most emergent effects.

The flexible nurb system (FlexiNurb) respondsquantitatively to different qualitative inputs (figure13.11). One of the goals of this innovation is tostandardize and automate technological processes andconvert them into techniques so that they can beperformed with a minimum of effort and provide stableplatforms from which further innovations can belaunched. The FlexiNurb process and the FlexiMolditself evolve simultaneously, and the production methodbecomes integral to the design of each component.Design and its implementation are no longer separatedfrom the conceptualization process and can be thoughtof as one continuous process of performativity.

Moving away from efficiency and optimizationtowards a model for design based on performativitytakes full advantage of techniques and their ability toinfluence behaviors and cultural endeavors. This not

13.13CatalyticFurnishings: twopossible variationswithin the flexiblemold system duringthe manufacturingprocess.

13.14CatalyticFurnishings: theform of one possiblevariation made offiberglass registersa hybridization ofseveral lifestyles.

Page 198: B. Kolarevic-Performative Architecture

191

only reveals greater interrelational potential betweentechnological, material and cultural lineages, butprofoundly increases their capacity to generate neweffects.

The scope and significance of such change ispotentially enormous. Technological choices give us away to bridge the gap between the technical and thecultural, immersing one within the other. The feedbackthat is developed through this immersion creates aplatform for innovation; the techniques that peoplegenerate through their use of technology not only exertpressure on technical refinement but enfold those

refinements within culture. Technological choicesdefine a world within which specific alternatives ofuses can emerge, and they define a subject whochooses among those alternatives. Fundamentaltechnological change is thus self-referential andtemporal. In the making of the world throughtechnology, we simultaneously enact great culturalchange. We need to operate within technology,developing technological practices that become partof technology as opposed to applying it to whateverwe are designing. Technological designing is afterall ontological designing.

13.15–16One variationof CatalyticFurnishingsbeing exhibitedat ArtistsSpace in NewYork.

Page 199: B. Kolarevic-Performative Architecture

192

NOTES1 John McKenzie, Perform or Else, New York: Routledge,2001, p. 6.2 Andrew Feenberg in Questioning Technology (London:Routledge, 1999) describes how the invariant elements of theconstitution of the technical subject and object are modified associally-specific contextualizing variables in the course of therealization of concrete technical actors, devices and systems.Thus technologies are not merely efficient devices, or efficiency-oriented practices, but include contexts as these are embodiedin design and social insertion.3 Larry A. Hickman, Philosophical Tools for TechnologicalCulture: Putting Pragmatism to Work, Bloomington, IN:Indiana University Press, 2001.4 A lineage is the evolutionary path demarcated by a single orcombination of cultural entities, through time, as the result ofreplication.5 Stephen Jay Gould, Bully for Brontosaurus, New York:Norton, 1991, p. 65.6 Georges Ifrah, The Universal History of Computing, London:Wiley, 2000, p. 239.7 Jeffrey Kipnis, Mood River, Columbus, OH: Wexner Centerfor the Arts, 2002, p. 53.

Page 200: B. Kolarevic-Performative Architecture

193

14

BRANKO KOLAREVIC

COMPUTING THEPERFORMATIVE

Page 201: B. Kolarevic-Performative Architecture

194

14

BRANKO KOLAREVIC

COMPUTING THEPERFORMATIVE

Page 202: B. Kolarevic-Performative Architecture

195

In avant-garde contemporary architectural design, variousdigital generative and production processes are opening upnew territories for conceptual, formal and tectonicexploration, articulating an architectural morphologyfocused on the emergent and adaptive properties of form.1

In a radical departure from centuries-old traditions andnorms of architectural design, digitally-generated formsare not designed or drawn as the conventionalunderstanding of these terms would have it, but they arecalculated by the chosen generative computational method.Instead of working on a parti, the designer constructs agenerative system of formal production, controls itsbehavior over time, and selects forms that emerge from itsoperation. The emphasis shifts from the “making of form”to the “finding of form,” which various digitally-basedgenerative techniques seem to bring about intentionally.

The new, speculative design work of the digital avant-garde, enabled by time-based modeling techniques, isprovoking an interesting debate about the possibilities andchallenges of the digital generation of form (i.e. the digitalmorphogenesis).2 There is an aspiration to manifestformally the invisible dynamic processes that are shapingthe physical context of architecture (figure 14.1), which, inturn, are driven by the socio-economic and cultural forceswithin a larger context. According to Greg Lynn, “thecontext of design becomes an active abstract space thatdirects from within a current of forces that can be storedas information in the shape of the form.”3 Formalcomplexity is often intentionally sought out, and thismorphological intentionality is what motivates theprocesses of construction, operation and selection.

This dynamic, time-driven shift in conceptualizationtechniques, however, should not be limited to the issues ofrepresentation, i.e. formal appearance, only. While we nowhave the means to visualize the dynamic forces that affectarchitecture by introducing the dimension of time into theprocesses of conceptualization, we can begin to qualifytheir effects and, in the case of certain technical aspects,begin to quantify them too. There is a range of digitalanalytical tools that can help designers assess certainperformative aspects of their projects, but none of themprovide dynamic generative capabilities yet.

PERFORMANCE-BASED DESIGNThe aesthetics of many projects of the digital avant-garde,however, are often sidetracking the critical discourse intothe more immediate territory of formal expression andaway from more fundamental possibilities that areopening up. Such possibilities include the emergence ofperformance-based design, in which building performancebecomes a guiding design principle, considered on a parwith or above form-making.

The current interest in building performance as adesign paradigm is largely due to the emergence ofsustainability as a defining socio-economic issue and tothe recent developments in technology and cultural theory.Within such an expansive context, building performancecan be defined very broadly, across multiple realms, fromfinancial, spatial, social and cultural to purely technical(structural, thermal, acoustical, etc.). The issues ofperformance (in all its multiple manifestations) areconsidered not in isolation or in some kind of linearprogression but simultaneously, and are engaged early onin the conceptual stages of the project, by relying on closecollaboration between the many parties involved in thedesign of a building. In such a highly “networked” designcontext, digital quantitative and qualitative performance-based simulations are used as a technological foundationfor a comprehensive new approach to the design of thebuilt environment.

It is important to note that performance-based designshould not be seen as simply a way of devising a set ofpractical solutions to a set of largely practical problems,i.e. it should not be reduced to some kind of neo-functionalist approach to architecture. The emphasisshifts to the processes of form generation based onperformative strategies of design that are grounded, atone end, in intangibilities such as cultural performanceand, at the other, in quantifiable and qualifiableperformative aspects of building design, such as structure,acoustics or environmental design. Determining thedifferent performative aspects in a particular project andreconciling often conflicting performance goals in acreative and effective way are some of the key challengesin performance-based design.

14.1The Dynaform BMW Pavilionat the IAA’01 Auto Show inFrankfurt, Germany (2000–01), architects BernhardFranken and ABB Architekten.

Page 203: B. Kolarevic-Performative Architecture

196

CALCULATING PERFORMANCE THENThe performative design thinking, framed by a broadlydefined performance agenda and supported by a rangeof digital performance analysis and simulation tools,as outlined briefly above, was envisioned decades ago.Back in the late 1960s and early 1970s, a group ofresearchers led by Thomas Maver at ABACUS(Architecture and Building Aids Computer UnitStrathclyde) at the University of Strathclyde’sDepartment of Architecture and Building Science,proposed that the building design be directly drivenand actively supported by a range of integrated“performance appraisal aids” running on computersystems.4

Digital building performance “appraisal aids” andperformance-based design were at the center ofcomputer-aided building design research for more thanthree decades — many of the essential concepts andtechniques were pioneered in the late 1960s and early1970s. For example, the first use of computer graphicsfor building appraisal was in 1966, the first integratedpackage for building performance appraisal appearedin 1972, the first computer-generated perspectivedrawings appeared in 1973, etc.5 The 1970s resultedin the “generation of a battery of computer aids forproviding the designer with evaluative feedback on hisdesign proposals,” enabling architects to “obtainhighly accurate predictions of such buildingperformance measures as heat loss, daylight contours,shadow projections and acoustic performance.”6

One of the first digital performance analysis toolsto emerge was PACE (Package for ArchitecturalComputer Evaluation), developed at ABACUS andintroduced in 1970 as a “computer-aided appraisal

facility for use at strategic stages in architecturaldesign,” which, unlike many of the efforts at the time,aimed “not on optimization of a single parameter buton production of a comprehensive and integrated set ofappraisal measures.”7 PACE was written in FORTRANand run on a time-sharing system; the “conversationalinteraction” was through a teletypewriter terminal. Theprogram measured costs, “spatial,” environmental and“activity” performance. The “spatial performance”component measured site utilization (plot ratio) andplan and mass compactness. Computing theenvironmental performance resulted in “plant sizeswhich [would] give adequate environmentalconditions,” while taking into account the heat gain andloss. The “activity performance” module measured “thedegree to which the relationships input under activityinformation are satisfied by the proposed scheme.”

The program would instruct the designer how tochange geometrical or constructional information, i.e.how to modify the design concept to improveperformance and then submit the modified design for“re-appraisal.” In the end, the “repetitive man/machineinteraction” would lead to “convergence of an‘optimum’ design solution.” A particularly interestingaspect of the program was its built-in capacity to“learn:” if the designer was satisfied with the scheme,the program would update the stored mean values usedin assessments.8

As is often the case with visionary ideas, much ofthe early work in digitally-driven performance-baseddesign was far ahead of its time both conceptually andtechnologically. But its time has now come, asperformance-based design is slowly but steadily comingto the forefront of architectural discourse.

14.2Finite-element analysis(FEA) stress analyses ofthe Dynaform BMWPavilion for the 2001Auto Show in Frankfurt,Germany, by Bollinger+ Grohman ConsultingEngineers, architectsBernhard Franken andABB Architekten.

14.3The FEA analysisof stresses for theSwiss Re building,London (1997–2004), by Arup,architect Fosterand Partners.

14.4The CFDanalysis of windflows for ProjectZED in London(1995) by Arup,architect FutureSystems.

Page 204: B. Kolarevic-Performative Architecture

197

SIMULATING PERFORMANCE NOWToday, digital quantitative and qualitative performance-based simulation represents the technological foundationof the emerging performative architecture describedearlier. Analytical computational techniques based onthe finite-element method (FEM), in which thegeometric model is divided into small, interconnectedmesh elements, are used to accurately perform

14.5An earlycomputerrendering of thestructural systemfor KunsthausGraz, Austria(2000–03),architects PeterCook and ColinFournier(spacelab.uk).

14.6The acousticalanalysis of thedebating chamberin the City Hall,London (1998–2002) by Arup,architect Fosterand Partners.

14.7Gaussian analysis,Experience MusicProject, Seattle(1999–2000),architect GehryPartners.

structural, energy and fluid dynamics analyses forbuildings of any formal complexity. These quantitativeevaluations of specific design propositions can bequalitatively assessed today thanks to improvements ingraphic output and visualization techniques (figures14.2–14.6). By superposing various analyticalevaluations, design alternatives could be compared withrelative simplicity to select a solution that offers desiredperformance.

Future Systems, a design firm from London, usedcomputational fluid dynamics (CFD) analysis in aparticularly interesting fashion in its Project ZED, thedesign of a multiple-use building in London (1995; figure14.4). The building was meant to be self-sufficient interms of its energy needs by incorporating photovoltaiccells in the louvers and a giant wind turbine placed in ahuge hole in its center. The curved form of the façade wasthus designed to minimize the impact of the wind at thebuilding’s perimeter and to channel it towards the turbineat the center. The CFD analysis was essential in improvingthe aerodynamic performance of the building envelope.

The original blobby shape of Peter Cook and ColinFournier’s competition winning entry for the KunsthausGraz, Austria (figure 14.5), was altered somewhat afterthe digital structural analysis by consulting engineersBollinger + Grohmann from Frankfurt revealed that itsstructural performance could be improved with minoradjustments in the overall form, by extracting theisoparametric curves for the envelope definition not fromthe underlying NURBS geometry but from the structuralanalysis. Likewise, Foster and Partners’ design for themain chamber of the London City Hall (figure 14.6) hadto undergo several significant changes after engineersfrom Arup analyzed its acoustical performance using in-house developed acoustic wave propagation simulationsoftware.

In Gehry’s office, Gaussian analysis is used todetermine the extent of curvature of different areas onthe surface of the building (figure 14.7). That way thedesigners can quickly assess the material performance,i.e. whether the material can be curved as intended, asthere are limits to how much a particular material with aparticular thickness can be deformed. More importantly,the curvature analysis provides quick, visual feedbackabout the overall cost of the building’s “skin,” as doubly-curved areas (shown in red) are much more expensive tomanufacture than the single-curved sections (shown ingreen and blue tones).

Page 205: B. Kolarevic-Performative Architecture

198

As these examples demonstrate, the feedback providedby visualization techniques in the current buildingperformance simulation software can be very effectivein design development. The software, however, operatesat the systemic level in the same passive fashion as twoor three decades ago. “Computer-aided appraisal” nowand back in 1980, as described by Thomas Maver, hasconsisted of four main elements: representation,measurement, evaluation and modification:

The designer generates a design hypothesis which isinput into the computer (representation); thecomputer software models the behaviour of thehypothesized design and outputs measures of costand performance on a number of relevant criteria(measurements); the designer (perhaps inconjunction with the client body) exercises his (ortheir) value judgement (evaluation) and decides onappropriate changes to the design hypothesis(modification).9

As noted by Maver, “if the representation andmeasurement modules of the design system can be setup and made available, the processes of evaluation andmodification take place dynamically within the designactivity as determinants of, and in response to, thepattern of explorative search,” which is a fairlyaccurate description of how performance analysis(“appraisal”) software is being used today.

CHALLENGESDesigning buildings that perform (i.e. “which work —economically, socially and technically”) is a centralchallenge for architects, as observed by Thomas Maverback in 1988.10 He called for the development of“software tools for the evaluation of the technical issueswhich are relevant at the conceptual stages, as opposedto the detailed stages, of design decision-making.”11

The challenges of developing such software,however, are far from trivial. Most of the commerciallyavailable building performance simulation software,whether for structural, lighting, acoustical, thermal or

air-flow analysis, requires high-resolution, i.e. detailed,modeling, which means that it is rarely used in conceptualdesign development. This shortcoming, and the lack ofusable “low-resolution” tools, is further compounded bythe expected degree of the user’s domain knowledge andskills. Another frequently encountered problem is thatcertain performance aspects can be analyzed in oneenvironment while other performative analyses must beperformed in some other software, often resulting insubstantial and redundant remodeling. Providing a certaindegree of representational integration across a range of“low-resolution” performance simulation tools is anecessary step for their more effective use in conceptualdesign.

Assuming that analytical and representationalintegration can be achieved, and that intuitive “low-resolution” performance simulation tools can be developed,additional challenges are presented by the need for activedesign space exploration. Instead of being used in apassive, “after-the-fact” fashion, i.e. after the buildingform has already been articulated, as is currently the case,analytical computation could be used to actively shape thebuildings in a dynamic fashion, in a way similar to howanimation software is used in contemporary architecture.12

In other words, the performance assessment has to begenerative and not only evaluative. For that to happen,however, a fundamental rethinking of how the digitalperformance simulation tools are conceptualized isrequired.

Ulrich Flemming and Ardeshir Mahdavi argued in1993 for the close “coupling” of form generation andperformance evaluation for use in conceptual design.13

Mahdavi developed an “open” simulation environmentcalled SEMPER, with a “multidirectional” approach tosimulation-based performance evaluation.14 According toMahdavi, SEMPER provides comprehensive performancemodeling based on first principles, “seamless and dynamiccommunication between the simulation models and anobject-oriented space-based design environment using thestructural homology of various domain representations,”and bi-directional inference through “preference-basedperformance-to-design mapping technology.”

Page 206: B. Kolarevic-Performative Architecture

199

PERFORMANCE-BASED GENERATIVE DESIGNAs Kristina Shea observed, “generating new forms whilealso having instantaneous feedback on their performancefrom different perspectives (space usage, structural,thermal, lighting, fabrication, etc.) would not only sparkthe imagination in terms of deriving new forms, butguide it towards forms that reflect rather than contradictreal design constraints.”15 As a structural engineer, shecites the form-finding techniques used in the design oftensile membrane structures (pioneered by Frei Otto) asthe nearest example of performance-driven architecturalform generation, in which the form of the membrane is

14.8Canopy designdeveloped usingeifForm for thecourtyard of theAcademie vanBouwkunst inAmsterdam (2002),designed by NealLeach, SpelaVidecnik (OFISArchitects), Jaroenvan Mechelen andKristina Shea.

dynamically affected by changing the forces that act onthe model. She notes that the form-finding techniques instructural engineering are generally limited to either puretensile or pure compression structures, and she promotesthe need for developing digital tools that can generatemixed-mode structural forms.16

According to Kristina Shea, a generative approach tostructural design requires a design representation of formand structure that encodes not only (parametric) geometrybut also a design topology based on the connectivity ofprimitives.17 The experimental software she developed,called eifForm, is based on a structural shape grammarthat can generate design topology and geometry, enablingthe transformation of form while simultaneouslymaintaining a meaningful structural system. Primitivesand their connectivity are added, removed and modifiedwith a built-in randomness in design generation, directedby a non-deterministic, non-monotonic search algorithmbased on an optimization technique called “simulatedannealing,” analogous to the “crystallization processes inthe treatment of metals.”18 The software develops theoverall form of a structure dynamically, in a time-basedfashion, “by repeatedly modifying an initial design withthe aim of improving a predefined measure ofperformance, which can take into account many differentfactors, such as structural efficiency, economy ofmaterials, member uniformity and even aesthetics, whileat the same time attempting to satisfy structuralfeasibility constrains.” The end product is a triangulatedpattern of individually-sized structural elements and joints(figures 14.8 and 14.9).

In a similar vein, I have proposed in a recent paper19

the development of generative tools based on performanceevaluation in which, for example, an already structuredbuilding topology, with a generic form, could be subjectedto dynamic, metamorphic transformation resulting fromthe computation of performance targets set at the outset.Such a dynamic range of performative possibilities wouldcontain at its one end an unoptimized solution and at theother an optimized condition (if it is computable), whichmight not be an acceptable proposition from an aestheticor some other point of view. In that case, a suboptimal

14.9eifForm:progressivegeneration ofthe canopydesign.

Page 207: B. Kolarevic-Performative Architecture

200

form, i.e. its discovery, in qualitative cognition. Even thoughthe technological context of design is thoroughly externalized,its arresting capacity remains internalized. The generativerole of the proposed digital techniques is accomplishedthrough the designer’s simultaneous interpretation andmanipulation of a computational construct (topologicalconfiguration subjected to particular performanceoptimizations) in a complex discourse that is continuouslyreconstituting itself — a “self-reflexive” discourse in whichgraphics actively shape the designer’s thinking process.

CONCLUSIONIn conclusion, the new “performative” approach to designrequires, at a purely instrumental level, yet-to-be-made digitaldesign tools that can provide dynamic processes of formationbased on specific performative aspects of design. There iscurrently an abundance of digital analytical tools that canhelp designers assess certain performative aspects of theirprojects post-facto, i.e. after an initial design is developed, butnone of them provide dynamic generative capabilities thatcould open up new territories for conceptual exploration inarchitectural design. More importantly, the emergence ofperformance-based generative design tools would lead to newsynergies between architecture and engineering in acollaborative quest to produce unimaginable built forms thatare multiply performative.

solution could be selected from the in-betweenperformative range, one that could potentially satisfyother non-quantifiable performative criteria.

This new kind of analytical software will preservethe topology of the proposed schematic design but willalter the geometry in response to optimizing a particularperformance criteria (acoustic, thermal, etc.). Forexample, if there is a particular geometric configurationcomprised of polygonal surfaces, the number of faces,edges and vertices would remain unchanged (i.e. thetopology does not change), but the shapes (i.e. thegeometry) will be adjusted (and some limits could beimposed in certain areas). The process of change couldbe animated, i.e. from the given condition to the optimalcondition, with the assumption that the designer couldfind one of the in-between conditions interesting andworth pursuing, even though it may not be the mostoptimal solution (figure 14.10).

In this scenario, the designer becomes an “editor” ofthe morphogenetic potentiality of the designed system,where the choice of emergent forms is driven largely bythe project’s quantifiable performance objectives and thedesigner’s aesthetic and plastic sensibilities. The capacityto generate “new” designs becomes highly dependent onthe designer’s perceptual and cognitive abilities, ascontinuous, dynamic processes ground the emergent

Page 208: B. Kolarevic-Performative Architecture

201

14.10An analysis ofsurface curvatureacross a range offormal alternativesextrapolated from acomputer animationby Matthew Herman(graduate student atthe University ofPennsylvania).

Page 209: B. Kolarevic-Performative Architecture

202

NOTES1 Branko Kolarevic, Chapter 2 “Digital Morphogenesis” in B.Kolarevic (ed.), Architecture in the Digital Age: Design andManufacturing, London: Spon Press, 2003, pp. 11–28.2 Ibid.3 Greg Lynn, Animate Form, New York: PrincetonArchitectural Press, 1999.4 Thomas W. Maver, “PACE 1: Computer Aided DesignAppraisal” in Architects Journal, July, 1971, pp. 207–214.5 Thomas W. Maver, “Predicting the Past, Remembering theFuture” in Proceedings of the SIGraDi 2002 Conference,Caracas, Venezuela: SIGraDi, 2002, pp. 2–3.6 Nigel Cross and Thomas W. Maver, “Computer Aids forDesign Participation” in Architectural Design, 53(5), 1973, p.274.7 Maver, “PACE 1,” op. cit.8 The program also offered eight perspective views of thescheme, which were drawn on a “graph plotter” driven by thepaper tape produced by the program. That was a“revolutionary” technological development back in 1970s!9 Thomas W. Maver, “Appraisal in Design” in Design Studies,1(3), 1980, pp. 160–165.10 Thomas W. Maver, “Software Tools for the TechnicalEvaluation of Design Alternatives” in Proceedings of CAADFutures ’87, Eindhoven, Netherlands, 1988, pp. 47–58.

11 Ibid.12 Kolarevic, op.cit.13 Ulrich Flemming and Ardeshir Mahdavi, “Simultaneous FormGeneration and Performance Evaluation: A ‘Two-Way’ InferenceApproach” in Proceedings of CAAD Futures ’93, Pittsburgh, USA,1993, pp. 161–173.14 A. Mahdavi, P. Mathew, S. Lee, R. Brahme, S. Kumar, G. Liu,R. Ries, and N. H. Wong, “On the Structure and Elements ofSEMPER, Design Computation: Collaboration, Reasoning,Pedagogy” in ACADIA 1996 Conference Proceedings, Tucson,USA, 1996, pp. 71–84.15 Kristina Shea, “Directed Randomness” in N. Leach, D. Turnbulland C. Williams (eds), Digital Tectonics, London: Wiley-Academy,2004, pp. 89–101.16 Ibid, p. 89.17 Ibid, p. 93.18 Ibid. Simulating annealing is described by Shea as “a stochasticoptimisation technique that tests a batch of semi-random changesgenerated by the structural shape grammar, measures theirperformance and then chooses one that is near the best. Theamount of deviation from the best is gradually reduced throughoutthe process, but not necessarily from one design to the next.”19 Branko Kolarevic, “Computing the Performative inArchitecture” in W. Dokonal (ed.), Digital Design, Proceedings ofthe ECAADE 2003 Conference, Graz, Austria, 2003.

Page 210: B. Kolarevic-Performative Architecture

203

15

BRANKO KOLAREVIC

TOWARDS THE

IN ARCHITECTUREPERFORMATIVE

Page 211: B. Kolarevic-Performative Architecture

204

15

BRANKO KOLAREVIC

TOWARDS THE

IN ARCHITECTUREPERFORMATIVE

Page 212: B. Kolarevic-Performative Architecture

205

We might distinguish between two kinds of spatialdisposition, effective and affective. In the first, onetries to insert movements, figures, stories, activitiesinto some larger organization that predates andsurvives them; the second, by contrast, seeks to releasefigures or movements from any such organization,allowing them to go off on unexpected paths or relateto one another in undetermined ways.

John Rajchman1

In the late 1950s, performance emerged in humanities— in linguistics and cultural anthropology in particular— and in other research fields as a fundamentalconcept of wide impact. It shifted the perception ofculture as a static collection of artifacts to a web ofinteractions, a dynamic network of intertwined,multilayered processes that contest fixity of form,structure, value or meaning. Social and culturalphenomena were seen as being constituted, shaped andtransformed by continuous, temporal processes definedby fluidity and mediation; thus a performativeapproach to contemporary culture emerged.

As a paradigm in architecture, performance can beunderstood in those terms as well; its origins can bealso traced to the social, technological and culturalmilieu of the mid-twentieth century. The utopiandesigns of the architectural avant-garde of the 1960sand early 1970s, such as Archigram’s “soft cities,”robotic metaphors and quasi-organic urban landscapes,offered images of fantasies based on mechanics andpop culture; they have particular resonance today, ascultural identity and spatial practice are beingrethought through performative acts that recode, shiftand transform meanings in a true, semiotic sense.

In this spirit, performative architecture can bedescribed as having a capacity to respond to changingsocial, cultural and technological conditions byperpetually reformatting itself as an index, as well as amediator of (or an interface to) emerging culturalpatterns.2 Its spatial program is not singular, fixed orstatic, but multiple, fluid and ambiguous, driven bytemporal dynamics of socio-economic, cultural and

technological shifts. In performative architecture,culture, technology and space form a complex, active webof connections, a network of interrelated constructs thataffect each other simultaneously and continually. Inperformative architecture, space unfolds in indeterminateways, in contrast to the fixity of predetermined,programmed actions, events and effects.

The description of performative architecture givenabove is one of many — its paradigmatic appeal liesprecisely in the multiplicity of meanings associated withthe performative in architecture.3 The increasing interestin performance as a design paradigm is largely due to therecent developments in technology and cultural theoryand the emergence of sustainability as a defining socio-economic issue. Framed within such expansive context,the performative architecture can indeed be defined verybroadly — its meaning spans multiple realms, fromfinancial, spatial, social and cultural to purely technical(structural, thermal, acoustical, etc.). In other words, theperformative in architecture is operative on many levels,beyond just the aesthetic or the utilitarian.

ARCHITECTURE AS PERFORMANCEAt the urban scale, architecture operates between theopposing poles of “smooth” urban space (by blending in)and urban landmarks (that stand out). Contemporaryavant-garde architecture advances the latter towardsarchitecture as performance art, which takes the urbansetting as a stage on which it literally and activelyperforms.

Some of the recent projects by Lars Spuybroek(NOX), such as the D-Tower4 in Doetinchem, theNetherlands (1998–2003), and Maison Folie5 in Lille,France (2001–04), can literally be seen as architecturalperformance pieces. D-Tower is a hybrid digital andmaterial construct (figure 15.1), which consists of abiomorphic built structure (the tower), a website and aquestionnaire that form an interactive system ofrelationships in which “the intensive (feelings, qualities)and the extensive (space, quantities) start exchangingroles, where human action, color, money, value, feelingsall become networked entities.”6 The complex surface of

Page 213: B. Kolarevic-Performative Architecture

206

the 12 m tower is made of epoxy panels shaped over CNC(computer numeric control)-milled molds (figure 15.2). Theepoxy monocoque shell is both the structure and the skin, andthus simultaneously multi-performative from the tectonic andbuilding physics perspectives (figure 15.3). The towerchanges its color depending on the prevailing emotional stateof the city’s residents, which is computed from responses ofthe city’s inhabitants to an online questionnaire7 about theirdaily emotions — hate, love, happiness and fear — and theseare mapped into four colors (green, red, blue and yellow),with a corresponding light illuminating the biomorphicsurfaces of the tower. The city’s “state of mind” is alsoaccessible through the website, which also shows the“emotional landscape” of the city’s neighborhoods. So,either by looking at the tower or the corresponding website,one can tell the dominant emotion of the day.8 The tower alsofeatures a capsule in which the city’s inhabitants could leavelove letters, flowers, etc. To motivate participation in thissocially and culturally performative urban and architecturalexperiment, a monetary prize of 10,000 euros is to beawarded to the “address with highest emotions.”

15.1D-Tower, Doetinchem,Netherlands (1998–2003), architect NOX/Lars Spuybroek.

15.2D-Tower:tectoniccomposition.

15.3D-Tower:structuralanalysis ofstresses.

Page 214: B. Kolarevic-Performative Architecture

207

In Maison Folie in Lille (figure 15.4), an old textilefactory that has been transformed into a new urban artcenter,9 the added multi-purpose hall (a black box)features an external, partially transparent skin, whoseintricate tectonic composition of metallic grilles producesvarying moiré patterns as one moves along it. Spuybroekrefers to this dynamic effect as a “static” movement, “ananimation of the vertical tectonics of the façade, …bending vertical lines in a complex pattern that produce awhole range of changes when walking or driving by,enhanced by the position of the sun.”10 There is also theliteral movement of changing lights placed behind themetallic grille of the façade, adding another layer ofintricacy to the building’s urban performance.

Dynamic display of light, i.e. changing light patterns,is a primary performative dimension in Peter Cook andColin Fournier’s Kunsthaus Graz, Austria (1999–2003;figure 15.5). BIX, the light and media installation designedby realities:united from Berlin, is inserted behind theacrylic glass layer to create a “communicative membrane”— a low-resolution computer-controlled skin, a “mediafaçade” that, through the display of signs, announcementsand images, hints at the activities within the building(figure 15.6). The performative aspects of the building areall geared towards an “urban communication strategy.”

The BIX light installation blurs the boundariesbetween the architecture and the performance medium; inthe Kunsthaus Graz “the medium is the message.”11

Extending McLuhan’s ideas to performative architecture,12

one could argue that mediated, animated architecturalskins have the potential to change how we relate to thebuilt environment and, reciprocally, how the builtenvironment relates to us, as manifested in MarkGoulthorpe’s Aegis Hyposurface project, described below.

Movement and performanceIt is often the movement of people around and through abuilding that gives architecture its performative capacity,as Maison Folie demonstrates. It is the experience ofarchitecture’s spatial presence and materiality — theengagement of the eye and the body — that makesarchitecture performative.

15.4Maison Folie,Lille, France(2001–04),architect NOX/Lars Spuybroek.

15.5Kunsthaus Graz,Austria (1999–2003), architectsPeter Cook andColin Fournier(spacelab.uk).

15.6BIX, the“communicativemembrane” forKunsthaus Graz,designersrealities:united.

Page 215: B. Kolarevic-Performative Architecture

208

In some recent projects, such as the Millennium Bridge inGateshead, UK (1997–2001; figures 15.7 and 15.8),designed by Wilkinson Eyre Architects, and theMilwaukee Art Museum (1994–2001; figures 15.9 and15.10), designed by Santiago Calatrava, the performativeis in the kinetic effects of architecture — it is not thesubject that moves but the object itself, creating anarchitecture of spectacle, an architecture of performance.

The Millennium Bridge in Gateshead — the “blinkingeye” bridge, as it is popularly called — is the world’s firstrotating bridge; the entire bridge rotates around pivots onboth sides of the river so that its tilt creates sufficientclearance for the ships to pass underneath (figure 15.9).The bridge’s elegant arches appear to trap movementeven when static; their dynamic metamorphosis has beendescribed as resembling the slow opening of a giant eyelid— hence the “blinking eye” moniker.

For the museum building in Milwaukee, SantiagoCalatrava designed a giant, movable wing-like sunscreen,a brise soleil, over a glass-enclosed reception hall. Madefrom fins ranging from 26 to 105 feet in length, theoperable brise soleil is raised and lowered to control theamount of light (and heat) that enters into the receptionarea (figure 15.10). Calatrava clearly designed theoperable brise soleil as an event, an urban performanceon Milwaukee’s waterfront. The performative, however, isnot limited to the kinetics of the sunscreen; there aremany “performances in geometry and engineering”13 in

15.7The MillenniumBridge in Gateshead,UK (1997–2001),architects WilkinsonEyre Architects,engineers Giffordand Partners.

15.8The MillenniumBridge: the bridge’sarches in the tiltedposition.

15.9The MilwaukeeArt Museum, USA(1994–2001),architect andengineer SantiagoCalatrava.

15.10The MilwaukeeArt Museum: thekinetic operationof the wing-likebrise soleil.

Page 216: B. Kolarevic-Performative Architecture

209

this building, as is the case with almost all ofCalatrava’s projects.

In addition to kinetic effects, a building’s skin canalso dynamically alter its shape in response to variousenvironmental influences, as the Aegis Hyposurfaceproject by Mark Goulthorpe shows. Developed initiallyas a competition entry for an interactive art piece tobe exhibited in the Birmingham Hippodrome Theatrefoyer, the Aegis Hyposurface is a digitally controlled,pneumatically driven, deformable rubber membranecovered with metal shingles (figure 15.11) that canchange its shape in response to electronic stimuliresulting from movement and changes in sound andlight levels in its environment, or throughparametrically-generated patterns. The dynamicperformance of the building’s skin can be either pre-programmed (determined) or in response toenvironmental changes (indeterminate, interactive).

The Bilbao effectIn these and previously discussed projects, architecture’surban performances aim beyond the spectacle of the kineticstructures, dynamic skins and the changing light patterns.From the stakeholders’ perspective (owners, municipal andregional governments, etc.), the intended performance of thosebuildings is primarily socio-economic; as urban landmarks,those buildings are meant to energize the urban contexts inwhich they are situated. By attracting the attention of localcity dwellers and global cultural tourists, they are seen as thesparks of urban and economic renewal. The performances (andoftentimes forms) of these buildings become highly politicized.

This political, socio-economic and cultural performativepotential of architecture is being rediscovered due, in largepart, to what is nowadays called the “Bilbao effect,” after thesocio-economic and cultural transformation of a sleepyprovincial town in northeastern Spain into a cosmopolitancultural magnet as a result of a bold architectural and culturalstrategy — the synergy of the global cultural brand of theGuggenheim Museum and the exuberance and expressivenessof Frank Gehry’s architecture.14 Not surprisingly, by reachingout for out-of-the-ordinary architectural tactics, citiesincreasingly expect miracles — hence, the curvaceous, light-animated forms of Kunsthaus Graz, the “blinking eye” bridgein Gateshead, and the wing-like museum in Milwaukee.

THE AESTHETICS AND ETHICSOF THE PERFORMATIVEAdmittedly, there is a considerable degree of novelty incomplex, curvilinear forms (in spite of numerous precedents)pursued with fervor by the contemporary architectural avant-garde. The strong visual and formal juxtapositions createdbetween “blobs” and “boxes” in traditional urban contexts, asis often the case, add to their “iconic” status and theirperception of being exceptional and marvelous. The expressiveform of the Kunsthaus Graz (figure 15.5), for example, is notaccidental — its performative intent is aimed at the socio-economic: by attracting people to the area, this “FriendlyAlien,” as the building is curiously named by its architects,with its strange, mediated skin, will act as a developmentcatalyst (aiming for the “Bilbao effect”).

15.11AegisHyposurface,architect MarkGoulthorpe/dECOi.

Page 217: B. Kolarevic-Performative Architecture

210

Appearance and performanceInterestingly, it is the surface — the building’s skin —and its complex morphology and tectonics, and notnecessarily the structure, that preoccupies the work ofthe contemporary (digital) avant-garde in itsexploration of new formal territories enabled by thelatest digital modeling software.15 On the other hand,Santiago Calatrava appears to reject the skin in manyof his projects and instead seeks to harness theexpressive powers of exposed structure for itsperformative potential, both literally, in the engineeringsense, and morphologically, for the beauty of force-driven formal articulation. Another strategy is to avoidthe binary choices of skin or structure and to reunifythe two by embedding or subsuming the structure intothe skin, as in semi-monocoque and monocoquestructures. The principal idea is to conflate thestructure and the skin into one element.

This search for performance in geometry andengineering, in turn, prompted a search for differenttectonics and “new” materials, such as high-temperature foams, rubbers, plastics and composites,which were, until recently, rarely used in the buildingindustry.16 For example, the functionally gradientpolymer composite materials offer a promise ofenclosures in which material variables can be optimizedfor local performance criteria, opening up entirely newmaterial and tectonic possibilities in architecture. Forexample, transparency can be modulated in a singlesurface, and structural performance can be modulatedby varying the quantity and pattern of reinforcementfibers, etc.17

From a historic perspective, balancingperformances in geometry and material is acontinuously present theme in architecture. Geometrywas often imposed onto the material, as manifested byvarious proportioning and other ordering systems. Adifferent approach was to let the geometry emerge fromthe material and its capacity to deal with compressionand tension (i.e. the material’s structural performance).Gilles Deleuze and Felix Guattari illustrate these twodifferent approaches with a brief reference to

Romanesque and Gothic architecture, where the latterrepresents a qualitative shift from the former, from the“static relation, form-matter” (Romanesque) to a “dynamicrelation, material-forces” (Gothic).18 As Deleuze andGuattari note, “it is the cutting of the stone that turns itinto material capable of holding and coordinating forces ofthrust, and of constructing higher and longer vaults.”19 Theforms “are ‘generated’ as ‘forces of thrust’ (poussées) bythe material, in a qualitative calculus of the optimum.”Such “Gothic” computation of form through a material wasa method, most famously, behind Antonio Gaudí’s work (hisinverted chain-link models) and projects by Frei Otto (theuse of soap bubbles, for example). In a contemporaryarchitectural scene, Lars Spuybroek’s “analog computing”of form, accomplished through the use of threads dippedinto liquids, is a direct antecedent of such a performative,materially-driven line of design thinking.20 For manydesigners in the contemporary architectural avant-garde,such as Mark Goulthorpe, Lars Spuybroek, BernhardFranken and others, the fluid synergies of form andmaterial, appearance and performance, architecture andengineering, are intrinsically embedded into the conceptualorigins of their work.

Environmental performanceAddressing the building’s appearance (“how it looks”) andits performance (“what it does”) increasingly requirescreating environmentally attuned buildings, whose physicalforms are shaped by environmental performances in respectto light, heat, energy, movement or sound. There iscurrently an interesting gap in the aesthetics (and ethics)between form-oriented or cultural performance-orienteddesigners (Frank Gehry, Greg Lynn, etc.) and those whosework aims at environmental performance (Thomas Herzog,Glenn Murcutt, etc.). On the other hand, there is anothergroup of designers — the ones whose work is neither tooformalist or environmentalist (Foster, Grimshaw, Piano,Sauerbruch and Hutton, Jourda and Perraudin, etc.). Thedesign strategies in the projects of the latter group varyconsiderably as they respond to different cultural andenvironmental contexts. In many of their projects, formaland environmental performative agendas were successfully

Page 218: B. Kolarevic-Performative Architecture

211

pursued in parallel. In the Swiss Re project in London(1997–2004) by Foster and Partners (figure 15.12), thedesign aims at maximizing the daylight and naturalventilation in order to substantially reduce (by half) theamount of energy the building needs for its operation. Thespiraling form of the atria at the perimeter, which runs theentire height of the building, is designed to generate pressuredifferentials that greatly assist the natural flow of air. Theaerodynamic, curvilinear form, besides affording acommanding, iconic presence, enables wind to flow smoothlyaround this high-rise building, minimizing wind loads on thestructure and cladding, and enabling the use of a moreefficient structure. In addition, the wind is not deflected tothe ground, as is common with rectilinear buildings, helpingto maintain pedestrian comfort at the base of the building.

It is interesting to note that many of the designersmentioned earlier — notably Norman Foster and NicholasGrimshaw, once labeled High-Tech and renamed Eco-Tech byCatherine Slessor21 — have explicitly stated their intentionsto improve the environmental performance of their oftenhighly visible buildings (figure 15.12). While one couldquestion the methodological consistency in their projects andwhether certain performative aspects, such as energyefficiency, were indeed maximized, these architects didmanage to consistently push the technological envelope ofenvironmental performance in their buildings.

An interesting example of a recent project that seems tocapture the broad agenda of performative architecture, fromcultural to environmental performance, is Renzo Piano’sTjibaou Cultural Center for the Kanak population of NewCaledonia (1991–98; figure 15.13). The “cases” that

15.12The Swiss Rebuilding in London(1997–2004),architect Fosterand Partners,engineer Arup.

15.13Section drawing ofthe Tjibaou CulturalCenter in Noumea,New Caledonia(1991–98),architect RenzoPiano, engineerArup.

15.15The solardiagram forthe City Hallbuilding.

15.14The City Hall inLondon (1998–2002), architectFoster and Partners,engineer Arup.

Page 219: B. Kolarevic-Performative Architecture

212

dominate the design, and that formally reference (butdo not imitate) Kanaks’ huts with their cone-likeshapes, were conceived with a particular culturalperformance in mind. The cones of the “cases” weretruncated for a more efficient environmentalperformance. The natural air flow within the buildingis then further enhanced using a system of computer-controlled louvers on the inner skin in “cases,” whichwas designed and developed through wind-tunneltesting and computer simulations by engineers at Arupand the Centre Scientifique et Technique du Batimentin France.

The performative design strategies can varyconsiderably as they respond to different contexts.Peter Cook and Colin Fournier’s Kunsthaus Graz(figure 15.5), which was discussed previously, featuresan expressive, biomorphic blobby form, and an acrylicglass “skin” whose primary function is to be a“communicative membrane” — a low-resolutioncomputer-controlled skin, a “media façade.”Interestingly enough, there is not a hint ofenvironmental performance in the Kunsthaus Grazproject, as if to suggest that the formal andenvironmental agendas are often incompatible —which cannot be farther from the truth. Foster andPartners’ City Hall in London (figure 15.14; 1998–2002), imbues an iconic, biomorphic form with a logicof environmental performance that calls for such aform in the first place. (The origin of the project waspurely formal — it attained its environmental logiclater in the development.) The “pebble-like” form ofthe building in the end resulted from optimization ofits energy performance by minimizing the surface areaexposed to direct sunlight. The building’s form is adeformed sphere, which has a 25% smaller surface

area than a cube of identical volume, resulting in reducedsolar heat gain and heat loss through the building’s skin(figure 15.15).

Foster’s performative approach to the design of theCity Hall building, for example, could imply a significantshift in how “blobby” forms are perceived. The sinuous,highly curvilinear forms could become not only anexpression of new aesthetics, or a particular cultural andsocio-economic moment born out of the digital revolution,but also an optimal formal expression for the newecological consciousness that calls for sustainable building.

CONCLUSIONSPerformative architecture is not a way of devising a set ofpractical solutions to a set of largely practical problems. Itis a “meta-narrative” with universal aims that aredependent on particular performance-related aspects ofeach project. Determining the different performativeaspects in a particular project and reconciling oftenconflicting performance goals in a creative and effectiveway are some of the key challenges in this approach toarchitecture.

In performative architecture, the emphasis shifts frombuilding’s appearances to processes of formation groundedin imagined performances, indeterminate patterns anddynamics of use, and poetics of spatial and temporalchange. The role of architects and engineers is less topredict, pre-program or represent the building’sperformances than it is to instigate, embed, diversify andmultiply their effects in material and in time.

The development of more performative techniques ofdesign is essential to this task. It necessitates a shift fromscenographic appearances to pragmatist imagination of howbuildings work, what they do, and what actions, events andeffects they might engender in time.

Page 220: B. Kolarevic-Performative Architecture

213

NOTES1 John Rajchman, Constructions, Cambridge, MA: The MITPress, 1998, p. 92.2 Performative architecture can also be seen as a generator ofnew cultural patterns. For example, organizers of a recently heldsymposium on performative architecture in Delft, theNetherlands (March 11, 2004), state that “instead of describingthe architectural object, performative architecture focuses onhow the architectural object and its process of productionperform by producing new effects that transform culture.” Formore details, see http://www.x-m-l.org/ and also http://www.lab-au.com/files/doc/performative_architecture.htm3 Performance is one of the most used (oftentimes misused andabused) but least defined concepts in architecture. As can begleaned from the chapters in this book, the ways in whichperformance is understood in architecture are oftencontradictory; the meanings associated with it are oftenarticulated as opposites.4 NOX (Lars Spuybroek with Pitupong Chaowakul, Chris Seung-woo Yoo and Norbert Palz) and Q. S. Serafijn, artist, and theV2_Lab (Simon de Bakker, Artem Baguinski), 1998–2003, aninteractive tower, a questionnaire and a website, for the city ofDoetinchem.5 NOX (Lars Spuybroek with Florent Rougemont, Chris Seung-Woo Yoo and Kris Mun), 2001, for the city of Lille — invitedcompetition (first prize). Model: Ouafa Messaoudi and EstelleDepaepe.6 From the NOX Architekten website: http://www.noxarch.com7 The questionnaire was written by the Rotterdam-based artist Q.S. Serafijn.8 Lars Spuybroek expressed his concern that the tower couldeasily end up showing only one color, presumably blue (forhappiness), given that Doetinchem is a Dutch city. He remarkedthat they may have to tweak the formula that computes the“total” emotion, so that the output is more varied. (The issue offinding appropriate “yardsticks” to measure qualitativeproperties that often defy quantification equally perplexes allperformative domains associated with the built environment,from social dynamics to environmental comfort.)

9 The complex of buildings contains exhibition spaces, artist-in-residence homes, clubs, Turkish baths, restaurants and soundstudios.10 NOX website, http://www.noxarch.com11 Marshall H. McLuhan, Understanding Media: The Extensionsof Man, New York: McGraw-Hill, 1965.12 According to McLuhan, technology effectively interferes withour senses and, in turn, affects the sensibilities of societies in whichwe live. That process, McLuhan argues, was and is still the causeof major cultural shifts. For more information see Eric McLuhanand Frank Zingrone (eds), Essential McLuhan, New York:BasicBooks, 1995.13 Rowan Moore, “INgeniUS” in Metropolis magazine, June2001.14 According to the Financial Times, in the first three years sinceits opening in 1997, the Guggenheim Museum in Bilbao has helpedto generate about $500 million in new economic activity, andabout $100 million in new taxes, as reported by Witold Rybzynskiin “The Bilbao Effect,” The Atlantic Monthly, September 2002.15 For more details, see Branko Kolarevic (ed.), “DigitalMorphogenesis” in Architecture in the Digital Age: Design andManufacturing, London: Spon Press, 2003, pp. 11–28.16 For more details, see Branko Kolarevic (ed.), “DigitalProduction” in Architecture in the Digital Age: Design andManufacturing, London: Spon Press, 2003, pp. 29–54.17 See Johan Bettum. “Skin Deep: Polymer Composite Materialsin Architecture” in Ali Rahim (ed.), AD Profile 155: ContemporaryTechniques in Architecture. London: Wiley Academy Editions,2002, pp. 72–76.18 Gilles Deleuze and Felix Guattari, A Thousand Plateaus:Capitalism and Schizophrenia, translated by Brian Massumi,Minneapolis, MN: University of Minnesota Press, 1987, p. 364.19 Ibid.20 For more details, please refer to Lars Spuybroek’s chapter inthis volume (Chapter 12).21 Catherine Slessor, Eco-Tech: Sustainable Architecture and HighTechnology, London: Thames and Hudson, 1998.

Page 221: B. Kolarevic-Performative Architecture

214

Page 222: B. Kolarevic-Performative Architecture

215

16

PETER McCLEARY

DIRECTION(AND A DIRECTOR)

PERFORMANCE

IN SEARCH OF(AND PERFORMERS):

Page 223: B. Kolarevic-Performative Architecture

216

16

PETER McCLEARY

DIRECTION(AND A DIRECTOR)

PERFORMANCE

IN SEARCH OF(AND PERFORMERS):

Page 224: B. Kolarevic-Performative Architecture

217

Existence is the extent of matter in space.Descartes1

What stands out, that is “existence,” from thehomogeneity of the world is different for each of thedisciplines that contribute to the production ofbuildings. They abstract different natural andhuman contexts from the total environment, attemptto satisfy different goals or performances, usedifferent theoretical models, appropriate differentmaterials to manifest their reality, imagine differentstates of equilibrium from open to closed or active topassive for their systems, engage different buildersand building processes to construct their designs,generate different flow patterns or vector diagramsto transmit their particular forms of energy, and usedifferent methods to synthesize their disparate viewsof the environment.

The production (design and construction) ofbuildings is a complex system. The presentdeconstruction of the system into many differentsubsystems of professions and crafts, theories andpractices, and materials and methods, illustratesthat ontogeny recapitulates phylogeny, i.e. today’sdesign and construction processes result from theirown historical development.

The difficult task is to combine the disparateperceptions of the performances to be satisfied, andto synthesize the geometry of the different flowpatterns (static and dynamic) into a coherent whole.

Possible future solutions might derive from HerbertSimon’s avoidance of optimized subsystems in favor of“satisficing” each subsystem through negotiation andcoordination,2 the hierarchical deconstruction of thewhole system, or change from a “state description” of theworld as sensed to a “process description” of the world asacted upon.

If “performative architecture” puts the satisfaction ofhuman equilibrium at the center of its purpose, a radicalchange needs to take place in the representation of thesystems of the architectural design and building process.

Many of today’s fashionable ideas were equallyfashionable in the 1960s; that is, the General SystemTheory, cybernetics, design methods, and mathematicaland physical biology. While today we revisit those ideas,in those earlier days calculations were constrained to logtables, then slide rules, and finally to electronic handcalculators. In 1961, in the structural division of OveArup in London, much time was spent inverting matrices(related to the geometry of the Sydney Opera House) witha hand calculator — they did not have a computer. Nowthat there are bigger, faster calculating devices, larger,more complex problems can be solved. In that distantpast, such problems could not be examined without theuse of a theory of hierarchies and methods of synthesis.

The concept of this chapter does not derive from myresearch at Penn in the 1970s with Robert Le Ricolaisand the rhetoric of tension, or with Louis Kahn and hissearch for essences. Thus, there is no mention of therheology of matter and structure and the isotropy ofspace. Rather, the idea is drawn from my 1960sprofessional experience with the Ove Arup BuildingGroup, later named Arup Associates.

Their key concept was that “the design andconstruction of buildings was the task of a multi-professional team.”3 This core idea was not a mereauxiliary hypothesis to other theories of architecture andengineering, but was sufficiently profound to deflect mystudies from the uniqueness of role of structure inbuilding to that of collaboration, coordination andtowards the belief that the design activity is polygamous(figure 16.1).

16.1The geometry ofmultiple grids forservices, spaceplanning andstructure — thenetworks are relatedbut not coincident.Mining andMetallurgyLaboratories,BirminghamUniversity, England(1965), architectand engineer Arup,London.

Page 225: B. Kolarevic-Performative Architecture

218

This chapter has three parts: a short prelude orintroductory section to its fugue, which is a polyphoniccomposition (like Arup Associates, with theirinterweaving of the logic of the many professionaldisciplines), and it concludes with a coda on thesynthesis of those logics.

PRELUDEIn 1951, ten years before I joined the Ove ArupBuilding Group, there was a conference and exhibitionin Darmstadt on the subject of “Man and Space.” Itserved as a commemoration of the 1901 Jugenstilexhibition on Behrens, Olbrich and others. The 1951exhibition exhibited photographs of the work of sometwentieth-century architectural masters: Taut, Loos, LeCorbusier, Gropius, Behrens and Olbrich again, FrankLloyd Wright, and others.

On the last day for the colloquium, or academicconference, Martin Heidegger presented his essay on“Building Dwelling Thinking,”4 and Jose Ortega yGasset spoke on “ The Myth of Man BeyondTechnique.”5 Among Heidegger’s hypotheses are thatbuilding has dwelling, that is, care, concern,cultivation, etc., as its goal; that dwelling and buildingare related as ends and means; and that to be humanmeans to dwell. His explication uses the case study of abridge. To an engineer, this is a revealinginterpretation of the nature of a bridge.

Ortega, often at odds with Heidegger — althoughboth were students of Edmund Husserl6 — argues thatdwelling does not precede building, and as man has notadapted to the world (i.e. he is an alien), he does notbelong, he needs a new world, thus he wants to build“like a lover without a beloved;” dwelling is not givento man, rather he fabricates it — and dwelling is aprivileged position or situation.

Apparently there were mixed reactions from thearchitects present at the colloquium, ranging from ageneral uneasiness to questioning the validity of theparticipation by philosophers.

This short prelude does not focus on this differencebetween the positions of Heidegger and Ortega, that is,

on the primacy or priority of building and dwelling. Instead,we note some similarities between their thoughts on thesubject of technology.

Not content with his contribution, in Darmstadt, to thediscourse on dwelling and building, Jose Ortega y Gasset(1883–1955) continued to explore the topic in his essays,leading to “Pragmatic Fields.”7 This well-known essay waspublished in Spanish after his death in 1955. In it he suggeststhat our lives consist in the articulation of many small worldsor territories (e.g. religion, knowledge, business, art, love),and that our life is nothing but a relentless dealing withthings. Properly speaking, in life there are no “things,” only inscientific abstraction do things exist, and realities that havenothing to do with us are just there, by themselves, andindependent from us. However, we must deal and occupyourselves with some things that are issues, that is, somethingthat must be done — a faciendum in Latin, or pragmata inGreek. He concludes that, therefore, we must contemplate ourlives as an articulation of “pragmatic fields.”

In his earlier 1939 essay, “Meditacion de la Tecnica” (or“Thoughts on Technology,” or “Man the Technician”8), Ortegasays that humans are ontological centaurs, half immersed innature, half transcending it: the extra-natural part is aprogram (or project) of life and that that program is limited toa pragmatic field. If we agree, we must conclude that, despiterumors to the contrary, we live in different worlds.

Much earlier, in his 1926 “Being and Time,”9 MartinHeidegger (1889–1976) had considered aspects of the topic ofthe 1951 conference. His “analysis of environmentality”discusses the spatiality of the “ready-to-hand” or the poeticsof use, in which we discover the world through touch andmaking things, and that this activity has its own kind ofknowledge.

The focus of this volume seems to be on the physics, notpoetics, of building. Heidegger, or Husserl, might say theemphasis is on the “present-at-hand.” He proposes thathumans encounter the world through equipment, and thatsince they use equipment in order to do something, it has thecharacter of closeness (or nearness), and further that thiscloseness of equipment has been given directionality. (See myessay on the transparency-opacity and amplification-reductionof equipment.10)

Page 226: B. Kolarevic-Performative Architecture

219

Heidegger continues that equipment has its place andthis place defines itself as the place of this equipment:

Thus the sun whose life and warmth are ineveryday use has its own places — sunrise, mid-day, sunset, midnight. The house has its sunny sideand its shady side; the way it is divided up intorooms is oriented towards these, and so is thearrangement within them, according to theircharacter as equipment. Thus we discover thespatiality of equipment. And space or geometricspace has been split up into places throughequipment.11

It is this similarity between Ortega and Heidegger, thatis, the existence of worlds that we experience throughmediation, that leads one to seek parallel hypotheses innatural philosophy, in addition to human, ontological ormetaphysical philosophy.

There are parallels in Ernst Mach’s “Space andGeometry,”12 where he correlates the separatephysiological spaces of vision, touch, smell, sound, andmovement, and in the idea of Descartes that existenceis the extension of matter in space.13

The conclusion to this prelude, drawn from ArupAssociates, Husserl, Heidegger, Ortega, Mach, andDescartes, is that each of us live in a differentabstracted world that is made apparent by our forms ofmediation (equipment, processes and theories), and thatwe use not only different means but also imaginedifferent ends.

FUGUEApplying this conclusion to the topic of “performativearchitecture,” we note that each time an aspect ofarchitecture, that represents a particular abstractedworld, is rendered measurable, there arises a newconcomitant specialty, whose authority is legalizedthrough licensure. What is common to all these specialdisciplines is that there is a continuum, real orimaginary, applicable to the whole architecturalenvironment and to the specialized or abstracted parts.

We will examine aspects of this continuum with itsenvironmental contexts (human and natural), goals(purposes and performance), theoretical models of eachdiscipline, the materials and equipment used, the buildingprocesses, their individual system geometries (or vectoranalyses), and the task of the synthesis of the separatesystems.

Environmental contextsFrom the human context, the architect considers relevantsocial structures (family, tribe, etc.), new agreements,institutions, human comfort, ergonomics, size andorientation of activities, movement of people, and socio-cultural preferences and values.

From the natural context, the structural engineerrecognizes the imposition of load actions due to gravity(self-weight and applied load), wind, snow, earthquake,ground movement and temperature changes.

Other specialists, often engineers, respond to changes inheat or coolness (skin), light (eye), sound (ear), smell (nose),air quality (lungs), and the flow, in both directions, ofliquids and gases (e.g. plumbing).

Apparently, they experience different worlds, or at leastthey are expected to examine different environments.

PerformanceEach of these professional realities seeks to satisfy differentends or purposes. The architect aims to optimize a numberof utility functions, that is, the product should be useful.Among those functions of what was formerly referred to asutilitas, is accommodation to the body (ergonomics) and itsactivities (space planning), or even to design spaces that“inspire” activities (Louis Kahn). Other utility functionsconsider human comfort and health (physiological),environmental stable equilibrium (homeostasis), safety andsecurity, durability, and the efficiency, life-cycle costs ofcapital, labor and energy.

The architect must also consider issues of appearance orgrace, once named venustatis. Here the appeal is to delightand pleasure, both at the level of individual preferences andcultural values. Sometimes the goal reaches beyond theconcerns of beauty towards the realm of the sublime.

Page 227: B. Kolarevic-Performative Architecture

220

A structural engineer examines, in part, issues ofstrength, once named firmitas. Here the expectationis for stability or no collapse, and strength or noyielding. To achieve these goals with efficiency, thetask is to minimize potential energy and to maximizestrain energy.

The specialist in light considers generalillumination and task lighting, but today rarelyconsiders metaphysical light. To perceive the space,this engineer must illuminate the objects to placeforms in relationship to each other.

The acoustics specialist considers the comfort ofhuman hearing.

And so forth.The different professions are responsible for

different environments; they experience differentworlds, and thus have different goals in mind. Theproblem remains to combine these differentperceptions.

Theoretical modelsEach specialist has a different theoretical modelwhose language ranges from verbal throughgraphical to mathematical. These theories are atdifferent stages of development in a continuum frommagic through craft (e.g. rules-of-thumb) andempiricism (e.g. analogies, codes of practice) toapplied science.

Most theoretical models in the building industryare closed systems from mechanics that focus on thestationary state. Today’s interest is in open or livingbiogenetic or biophysical systems in steady stateequilibrium. In the near future we will rediscover themuch fuller explication of systems that includeskinetics, dynamics and statics with its equilibria ofthe steady state, moving equilibria, and displacementof equilibrium. Perhaps even the principle of LeChatelier and Alfred J. Lotka’s 1924 text Elementsof Mathematical Biology14 will be rediscovered.There are many precursors to our present interest, inparticular, and to general systems. Meanwhile, eachspecialty uses theories unique to its discipline.

The architect models reality with a vector analysis of size,orientation and movement, flow of activities, concepts ofspace, such as classicism, modernism, deconstructivism,and emerging theory of complexity. Typology andmorphology offer palaeontological theories of space.

The structural engineer theorizes a mechanical worldof statics and dynamics, linear and non-linear, scientificanalogies, graphical methods, algebra, calculus,andmatrices. In G. H. Hardy’s essay, “ A Mathematician’sApology,”15 he says that matrices are beautiful becausethey are not useful in modeling physical problems. Whilewith Ove Arup, I tediously inverted matrices that modeledsome aspect of the geometry and structure of the SydneyOpera House. Is it possible that today there exist somebeautiful, useless ideas that in time will become useful? Asort of latent utility lies in the beautiful. Of course, thebalance between potential energy and strain energy is atthe core of the structural engineer’s thinking, as is theneed for upper and lower bound solutions. Problemsrelated to scale, or relative size, use the principle ofsimilitude or the theory of dimensions and itsunderstanding of the fundamental variables of mass,length and time.

The fact that many architects imagine the relationshipbetween structure and space in terms of mass aloneresults in architectural theories of structural behavior thatare appropriate only to heavy masonry structures. Mostengineers perceive space in a world of framed structuresrestrained by inertia, and some can visualize the space ofpre-stress and force.

The specialist on heat imagines a thermodynamicreality where the desire is for a “thermal steady stateacross time and thermal equilibrium across space.” Theirsis a world of forms of energy, microclimate, heat transfer,band width of radiation, heat that is sensible, latent,radiant or specific, and is transmitted by radiation,conduction or convection. In addition, their measures areof resistance and conductivity, thermal lags, solarradiation, mean temperatures and shading coefficients. Totheir perception, the presence of the sun is more evidentthan the pull of gravity. Heat is often aligned withhumidity and its theories of metabolism, absorptivity,

Page 228: B. Kolarevic-Performative Architecture

221

emissivity and charts that are either bioclimatic orpsychrometric.

Light and color have an array of theoreticalmodels derived from optics, with its rays that areabsorbed, reflected, refracted or transmitted.

Acousticians speak of sound pressure waves,frequency and wavelength, reflection, absorption,transmission, reverberation and resonance, pitchand tone.

Theories of the behavior of fire consider theimportance of flame spread, flammability andtoxicity.

Builders also have operative theories that dealwith the operations of man-machine complexes.These theories attempt to formulate a scientifictheory of action, sometimes named “praxiology.”16

Other models are drawn from operations research,decision theory and cybernetics.

Units of measureEach of the theoretical constructs conjures up its uniqueunits of measure, notwithstanding that the theory ofdimensions shows that all units are functions of mass,length or time. Whereas the architect might emphasizelength, area and volume, the structural engineermeasures in weight, stress, strain, modulus of elasticity,inertia and so forth. The other specialists focus on thedegree, thermal units per hour, joule and watt, candela,lumen, foot-candle, mass and decibel.

The many different languages, using differenttheories, words and units of measure, that are spoken inthe design and construction of buildings can lead tobabel, that is, a confusion of tongues.

MaterialsTo achieve its goals, each discipline uses its own specificbuilding materials. The architect might focus onmaterials that relate to perception of the space. Thestructural engineer uses many materials from the mostancient masonry and wood through steel and reinforcedor prestressed concrete to present-day carbon fiber. Inthe future, liquids and gases might replace thetransmission of forces through solids. Similarly, HVAC,lighting, acoustical and fire engineers, all use specializedmaterials to accommodate their different perceptions ofbuilding performance.

ProcessesDuring the processes of manufacture, fabrication andassembly, different subcontractors assemble each of thesedifferent materials, each using their own particular tools,processes and theories. Builders have their own mini-maxnotions of efficiency.

GeometriesEach profession articulates the geometry for efficientpatterns, layouts or types or system geometries for vectorflow, both its direction and magnitude, of the flux of themovement of people, air, stresses, waves (heat, light,sound, smell), liquids (water and sewage), and the rhythmor metric of the construction process (figure 16.2).

16.2Diagram illustratinga comparison ofsystems devised forscience buildings.Comparative systemsof buildings designedby Arup Associatesin Birmingham,Cambridge, Horshamand Loughborough,England; architectand engineer Arup,London.

Page 229: B. Kolarevic-Performative Architecture

222

While some architects are more interested in solidsthan voids, it is more usual for them to examine theflow through the voids, and describe its dimensions(linear, planar or volumetric), direction (one or twoway), its orientation to some reference datum,horizontal and vertical, the profile of the boundariesin terms of an inside surface (or intrados) and outsideboundary (or extrados), concavity or convexity. Todaywe see more use of complex geometrical shapes thatshows some knowledge of Gaussian curvature, e.g.torqued ellipses. Perhaps tomorrow we will see someuse of the seven elementary catastrophes, e.g. thehyperbolic umbilic. Similarly, there is an increasingawareness of the topology that models theconnectivity of spatial activities and their adjacencies.There remains the most basic understanding of flowas circulation through entrance, foyer (at rest),corridor (in motion), stairs (incline and algebra),ramps (incline and calculus), escalators (movingincline), elevators (vertical), and so forth.

The degree of isotropy, or the proportionalrelationships among the dimensions of space andtime, and topological measures are significantmeasures in the geometry of architectural space.

The structural engineer is more concerned withrheology, or the flow of stresses and deformations inthe solid matter. Similar to the voids of the architect,this geometry describes the solids in terms of theiraxes of restraint in one, two or three dimensions, andthe flow or “spanning” in one, two or threedirections. Various shapes are characterized as singlyor doubly curved, synclastic or anticlastic. Thefamiliar shapes are the cylinder, sphere, conic andhyperbolic sections, torus, and so forth. Rene Thom’sresearch on “chaos theory,”17 with its geometry ofseven “elementary catastrophes,” has broughtinterest to the shapes of the fold, cusp, swallowtail,butterfly, and three umbilic forms — hyperbolic,elliptic and parabolic. These complex shapes mustface the scrutiny of stress analysis and they may ormay not be as useful as minimum volume networks ofevolute and involute curves.

Studies in the isotropy of geometric space and therheology of matter show that space and matter aregeometric duals, that is, “images” of each other.

The expert on heating, cooling and air conditioningstudies the geometry of the flow in liquids, gases andwaves. They imagine open or closed channel flow in ductsor pipes. Among their systems are two-pipe, four-pipe,closed loop, central and peripheral, and single or multi-zone. These geometries allow more systematic andsystemic thinking than the more general typology of thehearth, foyer, inglenook, gazebo, porch, atrium andgreenhouse. Generally, the existence and use oftypologies indicates a formal logic in its infancy.Specialists in lighting and acoustics also have a set ofpreferred configurations that satisfy their performancecriteria. Builders too have their intuitive patterns andrhythms that derive from critical path methods,programmed enquiry research techniques and morerecent computer models that simulate the buildingprocess. Soon there will be a certified materials engineerwho alone knows the performance of materials.

Clearly, each profession lives in a different world,has different goals, constructs different theories, usesdifferent materials and methods of building, andimagines a different geometry.

CODA: SYNTHESIS ANDTHE INTERWEAVING OF COMPLEXITYTo synthesize these disparate perceptions we must lookdeeper than “design value analysis” and “economicevaluation methods,” such as payback period, life-cyclecost, return on investment, or comparative valueanalysis.

The production, both design and construction, ofbuildings is a complex system. The presentdeconstruction of that system into many differentsubsystems of professions and crafts, theories andpractices, and materials and methods, illustrates thatontogeny recapitulates phylogeny. In other words, theorganization of design and construction processes is theresult of their own fragmented perception and historicaldevelopment.

Page 230: B. Kolarevic-Performative Architecture

223

In this complex system, the flow path of eachoptimized subsystem interferes with the paths of theother systems (figure 16.3). In the 1970s, PhilipDowson, an architectural partner of the ArupBuilding Group, wrote that “as each professionoperates only within prescribed limits and carefullydefined boundaries of responsibility and authority,then an uncoordinated result must be expected,”and that “with five of the seven professionals withinan integrated design group having their traininglargely technically-based, it will be inevitable thatdesign solutions will be technically biased.”18

This suggests the need for team members fromother disciplines — perhaps from human-factors,sociologists, physiologists, behavioral and perceptualpsychologists, and philosophers — notwithstandingthe reservations of the architects at the 1951Darmstadt colloquium.

What methods might reconcile these seeminglycontradictory possibilities of optimized subsystems?The difficult task is to make a coherent whole byinterweaving the geometry of the multiple flowpatterns.18 Each specialist can describe severalalternative systems in terms of their geometric patternsand concomitant materials. The team, under theleadership of the architect, negotiates the interweavingof the geometries to where all the systems are satisfiedbut not optimized.19

As stated in the introduction, Herbert Simon’s“Sciences of the Artificial”20 describes a similarmethod that avoids optimizing subsystems by“satisficing” each subsystem through negotiation andcoordination. He also suggests benefits in a changefrom a “state description” of the world as sensed to a“process description” of the world as acted upon.

Cybernetics21 also suggests a theory of hierarchies22

to deconstruct the whole system. It proposes diagramsof immediate and ultimate effects, where “intenseinteraction implies spatial propinquity.”23

This is akin to Heidegger’s idea that humans “livein the space opened up by equipment.”

The above deconstruction of the whole system ofbuilding is similar to a medieval scholastic systemtheory24 in which they seek unity in diversity, that is,the unitas multiplex. The systemic relationship involvesa decomposition of the whole into parts, or explicatio,and the recomposition, or implicatio. The parts arerelated to the whole hierarchically, or concordantia.And, as in engineering aesthetics the logic should beseen, or manifestatio, and seen clearly, or claritas.

A temporary conclusion is that, if “performativearchitecture” puts the satisfaction of humanequilibrium at the center of its purpose, a radicalchange needs to take place in the representation of themultidisciplinary systems of the architectural designand construction process. Equally important is thatresearch must not be limited to the differentperceptions of the world, but must expand to includepredictive theories that synthesize these many differentperceptions.

16.3The geometry ofrelationships amongservices, spaceplanning, structureand partitions:LoughboroughUniversity ofTechnologyLaboratory,Loughborough,England (1965),architect andengineer Arup,London.

Page 231: B. Kolarevic-Performative Architecture

224

NOTES1 Stephen Gaukroger, Descartes: An Intellectual Biography,Oxford: Oxford University Press, 1997.2 Herbert A. Simon, The Sciences of the Artificial,Cambridge, MA: MIT Press, 1969.3 Philip Dowson, Architecture and Professional Integration,UK: Public Works Congress, 1972.4 Martin Heidegger, Poetry, Language, Thought: Building,Dwelling, Thinking, New York: Harper, 1971.5 Jose Ortega y Gasset, “The Myth of Man Beyond Technique”in Obras Completas, Madrid: Revista de Occidente, 1958.6 Edmund Husserl, Ideas: General Introduction to PurePhenomenology, London: Collier MacMillan, 1962.7 Jose Ortega y Gasset, “Pragmatic Fields” in ObrasCompletas, Madrid: Revista de Occidente, 1958.8 Jose Ortega y Gasset, History as a System: Man theTechnician, New York: W. W. Norton, 1962.9 Martin Heidegger, Being and Time, translated by J.MacQuarrie and E. Robinson, New York: Harper & Row,1962.10 Peter McCleary, “Some Characteristics of a New Conceptof Technology” in J. M. Stein and K. F. Spreckelmeyer (eds),Classic Readings in Architecture, Boston, MA: McGraw-Hill,1999, pp. 169–179.11 Heidegger, Being and Time, op. cit.12 Ernst Mach, Space and Geometry: In the Light ofPhysiological, Psychological and Physical Inquiry, Chicago,IL: The Open Court Publishing Co., 1906.

13 Rene Descartes, “Principia Philosophiae” in Charles Adam andPaul Tannery (eds) Oeuvres de Descartes, 13 volumes, Paris:Leopold Cerf, 1897–1913.14 Alfred J. Lotka, Elements of Mathematical Biology, New York:Dover Pub., Inc., 1956.15 G. H. Hardy, A Mathematician’s Apology, Cambridge:Cambridge University Press, 1967.16 Tadeusz Kotarbinski, Praxiology: An Introduction to the Scienceof Efficient Action, New York: Pergamon Press, 1965.17 Rene Thom, Structural Stability and Morphogenesis, Reading:Benjamin, 1975.18 Derek Sugden, “Planning for the Unknown” in BritishConference on Steel in Architecture, November 1969, pp. 157–167.19 Philip Dowson, “Building for Science” in Architectural Design,April, 1967, pp. 160–170.20 Simon, op. cit.21 Norbert Weiner, Cybernetics: On Control and Communication inthe Animal and the Machine. Cambridge, MA: MIT Press, 1965.22 Howard H. Pattee, Hierarchy Theory: The Challenge of ComplexSystems, New York: Braziller, 1973.23 W. Ross Ashby, Design for a Brain, London: Chapman and Hall,1966; W. Ross Ashby, An Introduction to Cybernetics, London:Methuen, 1968.24 Erwin Panofsky, Gothic Architecture and Scholasticism, NewYork: Meridian, 1957; Maurice De Wulf, The System of ThomasAquinas, New York: Dover Pub., Inc., 1959; Umberto Eco, TheAesthetics of Thomas Aquinas, translated by Hugh Bredin,Cambridge: Harvard University Press, 1988.

Page 232: B. Kolarevic-Performative Architecture

225

17CONCEPTUALPERFORMATIVITY

WILLIAM BRAHAMHARALD KLOFTDAVID LEATHERBARROWALI RAHIMMAHADEV RAMANANDREW WHALLEYBRANKO KOLAREVICALI M. MALKAWI

Page 233: B. Kolarevic-Performative Architecture

226

17CONCEPTUALPERFORMATIVITY

WILLIAM BRAHAMHARALD KLOFTDAVID LEATHERBARROWALI RAHIMMAHADEV RAMANANDREW WHALLEYBRANKO KOLAREVICALI M. MALKAWI

Page 234: B. Kolarevic-Performative Architecture

227

KOLAREVIC: We have heard very different versions orvisions of what performativity or performance meansin architecture; as we have said at the outset, such anoutcome was expected. The challenge is to give somekind of overall coherence to the notion ofperformativity or performance in architecture.

In the two panel discussions we will examine athematic territory that lies between what we see as twopoles of performativity. At one end are the conceptualdomain and the performativity of the virtual. At theother are the operative domain and the performativityof the real. We hope to situate the discussions betweenthese two poles — the conceptual realm and theoperative realm — perhaps getting a little closer to thefirst one in this discussion, and to the second one in thediscussion that follows. We are not really constrainingthis discussion to the conceptual realm as, in my view,it is impossible to separate the two.

We will begin the discussion with a few questionsand issues associated with the conceptual realm. Towhat extent do the issues of performance, broadlyunderstood as we have heard, figure in conceptualdesign? Is there such a thing as performative form-making or performative space-making? What is therelationship between performance simulation andform-making, and how can one inform the other andvice versa?

So, the first question is whether there is such athing as performative space-making or performativeform-making?

WHALLEY: I saw David Leatherbarrow’s definition ofperformance in architecture as having two halves, andI was actually trying to work out in which of the twoour work would be located. Undoubtedly, our earlywork sits in the first half, which is highly programmed,highly functioning, program-specific spaces. As ourwork has evolved, it was clearly moving and blendingtowards the second.

In the early days we were really quite obsessedwith function, to the exclusion of all other things. Forinstance, the early factories were entirely about

designing a skin and that was it. To some degree suchdisposition also had to do with the type of architecture wewere given at the time, which was fairly simple architecture,low budget, so the outcome was a one-liner.

As projects became more sophisticated and complex, ourwork started to take on broader aspects, in addition to justbeing purely functional. In other words, at the end of the daywe actually stepped back and asked: but is it beautiful? Isthere a materiality to it? How will it weather? How will itlook in twenty years’ time? How will it sit in thisenvironment? How will people relate to it? Is it the righttactile material on the inside?

There is a whole series of filters as well as the pureprogrammatic. To some degree, high-quality architecture hasto take both into account. Architecture has to beprogrammatic, it has to be functional, it has to deliver whatpeople want it to do. One cannot just exclude all of that.

To some degree, David Leatherbarrow was showingarchitecture that was just architecture in its own right andthen took on a programmatic quality, in addition to its sheerbeauty and tactile quality. For me, architecture has to movetowards that programmatic quality …

KOLAREVIC: It has to perform?

WHALLEY: It has to perform, but at the same time it has toalso take on these other levels. One cannot just work at onelevel.

KOLAREVIC: So, would you say that Grimshaw is engagedin performative space-making?

WHALLEY: Well, yes. As I explained, our work is process-driven; it is not stylistically predictable — it is very muchdriven by the individual program of a particular project.Through that process, we take on many other architecturalissues, such as the issues that have to do with the art ofarchitecture. When we used the water wall to cool theBritish Pavilion building, we designed a functioning systemto demonstrate how it could work. But the project is actuallyabout the senses, about creating the feeling of an oasis of acool space in a hot climate, of stepping through the sound of

Page 235: B. Kolarevic-Performative Architecture

228

water, stepping through into a pool, gliding over thatpool and seeing the shimmering water going over theglass.

So, yes, our architecture is performative … Itperforms, and it also works in an architectural way, anartistic way, in a sensuous way. Good architectureactually has to address both.

KOLAREVIC: It has to perform in both ways.

WHALLEY: Yes.

RAMAN: There are some fairly simple examples wherethe connection is quite easy to see. The design of asuspension bridge aims for something very beautifuland elegant but, in fact, every part of it is optimizedfor performance. The design of a concert hall is similarin its aims; for example, take the Kimmel Center forPerforming Arts in Philadelphia and the interactionsthat went on between the architect who wanted toachieve a certain type of space and the acousticianwho wanted to achieve a certain acoustic for theperformance, and the theater consultants andmechanical engineers and so on. There was a processthere that is simultaneously giving form and dealingwith performance issues because ultimately theperformance of that space is very, very critical. Thereare many other places where the definition ofperformance is a little vague and that is whensometimes we would go astray.

RAHIM: If you think about performance andperformativity in terms of architectural space, youbegin to think about cultural efficacy, organizationalefficiency or maybe technological effectiveness. This isa really important issue and I think we have all beengrappling with it in a disparate manner.

If you think about performance, it really dependson who you speak to with regards to how it has beenperceived and how it can be influential on form-making. If you begin to understand each condition onits own and, in fact, if you ask any technical person

how they define performance, they would define it in the waythey do things; for example, it is 80% cooling efficiency ifyou are talking to a mechanical engineer, and the engine is37% fuel efficient, etc. Then they follow it up with a verygeneral definition of performance that is not quite thatclear. Each of their definitions is localized to them.

Mahadev Raman, for example, places performance intoa context specifically related to organizational efficiency. Bymoving away from the machine model into a networkstrategy, conditions such as quantities become qualities.

If we read performance with regards to form-making, itis really about intensity. As William Braham has mentioned,it is about an intensity of unseen vectors that are in themilieu, that then formulate and, at particular moments intime, become intense and produce an effect. It is reallyabout intensifying the conditions so that there is a potentialof producing an effect. If you zoom out and if you see thewhole world with regards to performance, it is really abouta cloud of contested vectors with particular intensitiesemerging. That is the only way to cut across the differentparadigms of performance, shifting from the technical tothe technological which is cultural.

KOLAREVIC: Ali Rahim talked about the notion of dynamicmultiplicity. So if I could reduce, using the word “reduce”here operatively, the notions of performance to its essence, itis really about this dynamic multiplicities that are present,is it not?

LEATHERBARROW: That is related to what AndrewWhalley said about levels and the simultaneity of levels. Iwant to add perhaps a slightly separate issue. It seems partof what we have been struggling with is what yardsticks wehave, what measures we bring to bear on the several kindsor levels of performance one expects of the building, leavingaside for a minute the problem of form-making or of design.Do we have a yardstick for beauty? Certainly not. Do wehave one for health — a thermometer?

The concept of quality seems to plunge us into greatuncertainty about measure. Part of the issue here is thespectrum of measuring tools we are willing to tolerate, fromthe most discerning and objective to those that are rather

Page 236: B. Kolarevic-Performative Architecture

229

more personal and subjective. At some point thearchitect’s authority shades off into that of the client,the constituency, the public at large. This happenswhen one says this room performs more or less the waywe expect. In reaching such a conclusion, each of usinvokes our understanding of common culturalexpectations.

To quote Ortega y Gasset, the good architect is theone who knows more than an architect knows; thearchitect remembers what it is like to be in a lectureroom, and the standard, the measure, the criterion ofjudgment, is not a matter of professional expertise butrather cultural knowledge. If we want to marginalizeall of that, and class as performative those phenomenaof architecture that can be measured according to therulers of our expertise, I think we are reducingarchitecture.

We have to develop a concept of performativitythat is sufficiently nuanced and subtle to embracedifferent kinds of measure. But I would not absorb alljudgments of quality, intensity or beauty into what canbe measured objectively, nor would I say that it is aseparate issue from instrumentality. We should haveconcepts of instrumentality and performativity thatsustain these different kinds of reading or aspects ofmeasurement. That is where a number of the issuesaround prediction, accuracy, knowledge and foresightbecome rather difficult.

KLOFT: Other industries would run into problems ifthey could not measure formal, aesthetical and otheraspects of what they create as consumer products.That is the difference between creating the productand what we are doing — the unique solution. Otherindustries establish mainstream rules as to what worksright, so they do marketing, they create a world for theproduct — which we do not do or cannot do for abuilding.

Performance is not the same as efficiency;performance is much more than optimization. Whatwe did in our freeform projects was not about havingan optimized process.

KOLAREVIC: In my view, we cannot separate aesthetics fromperformance in architecture; the two are inseparable in manyways. We should not reduce the issues of performance to thethings that can be quantified; it is much more interesting todiscuss the intangibilities that begin to qualify performance insocio-economic and cultural terms.

We should perhaps frame the discourse of performance ina temporal fashion, i.e. what presently defines theperformative in architecture; in other words, what is acceptedas a structure, as architecture that performs well today, notonly in a technical sense, in the sense of building physics, butalso in a socio-economic and cultural sense.

What is performative today may not be so ten years fromnow, which goes back to what David Leatherbarrow hasdefined as eventmental architecture. When we talk about theevents, however, we cannot really divorce thinking of the“unscripted” events from what is scripted in architecture —the program that the architects are typically given — becausethe program assumes or attempts to predict certain events.How do our practicing colleagues deal with thesecontradictions of scripted versus unscripted?

LEATHERBARROW: Ali Rahim said that part of the task ofdesign is to recognize first what is inevitable and, second, thevalue of things over which one does not have control. Thechallenge is to see both as part of, not added to the process —both equally internal.

Once I had the pleasure of hearing Yehudi Menuhinperform in Teatro Olympico. It was a perfect instrument, itperformed perfectly. But the genius of the performance wasMenuhin’s ability to modify it to the room. Speakingconceptually or epistemologically, one faculty the architectmust possess is foresight — this technology will lead to thatresult. My counter position or complimentary position is thatthe architect must also possess ingenuity. In this particularcircumstance, the performance has to be adjusted; you playsofter or more loudly, more rapidly or more slowly. In otherwords, you have the technical procedure, the optimal solution,but the architectural insight says in this circumstance it canonly be done this way, and that is part of the project; it is notadded to it, it is not secondary to the initial so-calledobjective, repeatable and instrumental process.

Page 237: B. Kolarevic-Performative Architecture

230

Within the instrumentality of architectural work theremust be adjustment and internal correction. Thearchitect’s acuity and judgment must be brought to bearon the particularity of the case. Ingenuity is alwayscontingent, never planned, unpredictable, and calls for acertain spontaneity of the design. That is where I thinkthe building or the design becomes performative in amuch larger sense.

The concept of script or plan, which is what we call itin design, is inadequate to this second level ofunderstanding, which is really when the project takes off,when it is more than what could have been done in anycircumstance, because it is only possible and uniquelyrelevant in this one. When this happens the performancecomes alive, the building starts to live and breathe withthe client, with the site, with the time and the context, ina way that one is really excited about, as opposed toconfident in. So one could put confidence at risk for thesake of relevance, immediacy, concreteness andengagement, and that is what I mean by unscripted, theparticularity of the case.

BRAHAM: We are essentially using different languagesto describe the same thing, which is the ability to say withsome certainty what will happen within the dimension of,for example, producing a temperature in a building. Andyet, there are much larger conditions in which, fromevery conceivable different position, that does or does notmean something, is or is not good; and there is confusionbetween that and much more narrow activity.

The difference emerges when we talk about structuralor constructional performativity, which has to do with thedurability of the artifact, and the other dimension ofenvironmental performativity, which has completely to dowith people’s experience (maybe even not within theartifact). That difference manifests itself in teaching, astwo different trainings in backgrounds.

It is a much more profound difference than just aninstitutionalized one. It is quite a different thing to talkabout the object and its experience. It is far, far moredangerous when we start thinking that by adjusting theexperience or the conditions of experience that we really

have some effect over the people that come into a building.It is perhaps our hubris that we want to make people happysomehow when they come into our buildings.

WHALLEY: But that probably minimizes the problem to acertain degree. We can say we will design for people;perhaps one way of measuring the performance of a buildingis to try to understand the subjective interaction between thebuilding and all people who are interacting with it. Thatmight lead us to different things.

BRAHAM: I was thinking about Mahadev Raman’s remarkthat commissioning actually made more of a difference insome cases than design.

WHALLEY: But the missing piece in all of that is theassumption that we can actually quantify precisely theexperience of, for example, thermal comfort. Many papershave been written on this and there are design guidelinesthat have come to be accepted, but I know that when I amfeeling a little ill the condition that I would find thermallycomfortable is quite different to when I am feeling full oflife and energy. So even when there are conditions that canbe measured, the subjective response or the effect on thereceiver of those conditions is not measurable in anypersonal aspect.

BRAHAM: The other connection I would posit is thatperformativity occurs with some kind of feedback betweenthe designer and the milieu or the object and the subject.Although these notions of feedback are a much overworkedconcept, if understood in a much broader sense, it actuallyhelps us somewhat out of the dilemma of trying tooverextend our reach and understand the degree to whichthese things are constantly being shaped and made, both bythe unpredictable events and by the ones that we haveimagined will be accommodated.

KOLAREVIC: I want to ask our speakers to be specificabout what has challenged their notion of performativearchitecture — in other words, was there something thatyou heard that you disagreed with?

Page 238: B. Kolarevic-Performative Architecture

231

MALKAWI: Let me clarify that a little bit. When westarted, Branko Kolarevic had a certain idea aboutperformance, and I did too. We could have hadprobably two separate conferences, but we thoughtthe most important and interesting thing would be toexplore the integration between the two. One of ourgoals is to try to understand the apparent“disconnect” between these two worlds, the scientificand the artistic, and the influence that computationhas on bringing them closer together.

BRAHAM: What did you disagree on?

KOLAREVIC: The one thing that I find annoying isthe engineers’ love for the things that can bemeasured and quantified, and the designers’ aversionto the things that can be quantified and measured.Some of our speakers alluded to this tension thatexists between things that can be quantified andqualified, and things that are simply intangible. Tome, this tension between these two poles is a veryinteresting territory to address and explore, and I amcurious to hear what your positions are.

KLOFT: For some engineers, it is not so interesting towork on performative architecture because it isalways optimized; it is much more interesting to workon a performative process and do some really extremethings, like we did, and bring techniques forward. Soit is not only architecture that should be performativebut the process should be too.

WHALLEY: I was really impressed by Harald Kloft’sattitude in working with Bernhard Franken; you aregiven a form by Bernhard Franken and he says youcannot change it, you cannot touch it. It goescontrary to what you have been trained to do as anengineer — you are trained to optimize things, but hesays: “I don’t really need an optimized structure, Ijust want the form and make sure that it works thatway.” I would die if I were an engineer under theseconditions …

KLOFT: No, no … you develop a new technique, a differentfocus. For example, on the “Bubble” project weconcentrated on the realization, how to form the acrylicform, etc. It was very good for us as engineers to have afixed geometry from the beginning because if it changedevery week then we could not focus on how to deal withnew techniques of realization.

KOLAREVIC: Fixing things in the process was necessary,was it not?

LEATHERBARROW: We have to put an end to thisbusiness of qualitative and quantitative; two worlds, blackand white, good and bad — forget it, it is not true.

KOLAREVIC: I never said that either one is good or bad.

LEATHERBARROW: They are not two, they are one — itis one spectrum of decisions, more or less certain. How dowe judge whether or not this lecture room is too bright ortoo dark, too warm or too cool, its construction a waste ofmaterials or not?

We make these judgments because we have been inlectures before; we have been in rooms like this, and eachof us carries with us a whole history of culturalbackground, of experiencing lectures. There could be ameasure or a spectrum of measures, some very precise,some reasonably so, some barely at all, and design workswith all those levels. Nobody prefers quantitative orqualitative; design intelligence knows which kinds of thingsrequire which kinds of measure. The profound mistake is totry to apply this kind of measure to that kind of thing, as ifthere could be a yardstick of beauty. But that does notmean to say we could not look at this room and say it ispretty ugly or it is marvelously beautiful; one could saythat there would never be any agreement about that, but Iwould say there would be more or less disagreement. If youcould tolerate degrees of certainty, and if you could seedesign wisdom as knowing which kind of problem requireswhich kind of certainty, then I think you can avoid thispolarity, which I believe we should do. If we want to talkabout the performance of a space, we should see it as

Page 239: B. Kolarevic-Performative Architecture

232

embracing that full spectrum of measures. It is justthat some are quantified in one way and othersquantified in a different way.

So if it is true that there is something like acultural norm that is called the lecture theater — andwe all basically know what it is (you sit there and westand here, that is darker, this is lighter) — we can saythis room is more or less correct with respect to thatnorm. Is it fixed? No, it has to change in each case,and those changes give design its freedom. My viewhere is that you are completely wrong when youpolarize qualitative and quantitative.

KOLAREVIC: I perfectly agree with what you say. Iam not arguing for either of the poles, and actually Iam not separating the qualitative from the quantitative.I am pondering the worlds of engineering and theworlds of architecture, and how they are commonlyperceived and what they stand for. Architects love tothink that they deal with the intangibles and theengineers love to think that they are dealing with thethings that can be quantified and therefore qualified.Those are the polarities that I am trying to establish.

WHALLEY: Engineers like to work in vectors, do theynot? A scientist likes to do an experiment and if theexperiment produces the same result ninety times thenthat is a success, whereas artists and architects work incircles; we can design something a hundred times thatappears to be the same, yet to the artist or the designerit is a different exploration each time. These are thetwo poles that we are trying to blend in architecture.

LEATHERBARROW: You have the scientist andperhaps the artist at the two poles, you have thearchitect and the engineer moving substantially in fromthose poles into a middle ground where you cannot beas extreme as you implied in your statement.

KOLAREVIC: The design practices that manage tosomehow operate in this middle ground between thetwo extremes are the ones that are successful in the

contemporary context. Grimshaw is a practice that occupiesthat middle ground between the polarities I have described.Harald Kloft’s engineering practice is willing to accept thisfuzzy ground that architects have now offered them as agiven — without questioning?

KLOFT: We are not interested in measuring things. Our aimis to support the architecture, to provide a performativeprocess for a performative architecture.

Equally important is the issue of adaptability —buildings cannot change like other objects. In the future, weshould look more into this issue of change, i.e. how we candeal with the vectoring, the climate concepts.

LEATHERBARROW: I disagree, buildings do change.

WHALLEY: Architecture that was traditionally quite staticis becoming more and more active, adapting itself to theseasons and to change. You can have an environment where,when the summer comes, you can allow the air temperatureto go up a bit, but you brought in the shading and thesunlight permeates through that shutter. That speaks to thequalitative — it is a sensuous thing. You are in touch withthe season. The building is performing in a certain way.

I suppose the antithesis to this kind of high performance— performative — architecture would be a box where youswitch on the air conditioning and set it at 20°C and it holdsit winter to summer, which is what has happened in the1950s and 1960s. Whereas now, architecture is performingin interesting ways. It uses architectural devices in the sameway a boat can lift up the sails and capture the energy of thewind. Architecture now responds in similar ways, as one cansee it in traditional vernacular buildings. It always has hadthat location-based response to where it was placed. Thatidea of placement in architecture has to do with the placeand the season, and so on. That is when architecture reallybecomes a rich experience.

RAMAN: The approach previously used was a kind ofobjective performativity, where you are trying to achievecertain measurable goals, or certain styles, or certain results.The new approach to designing environments is much more

Page 240: B. Kolarevic-Performative Architecture

233

of a subjective performativity. So, if you say that theoccupants ought to be able to adjust the shading, thelight level, the temperature, etc., to suit whatever theirparticular mood or requirements happen to be at thetime, then in a sense that becomes a little more preciseas a measure of performativity that you get if youachieve that flexibility in the space.

LEATHERBARROW: Absolutely. That kind ofattunement is more precise than the optimization whichis indifferent to the particularity of the circumstance.

According to Cicero, a good public speaker is theone who can adjust the speech to what the audience iscapable of understanding. This judgment of what isright in these circumstances is not subjective — it issituational performativity. In particular situations,performances unfold in particular ways. When thecontext varies, the performance cannot remain thesame. It will be ineffective. So the challenge is tooptimize in a given context. It is not optimal in itself assuch in any possible application; rather, the realarchitectural optimization is a situated performativity.

Architecture must engage the contingencies as partof its rationality — a contingent rationality, a weakrationality, a situated performativity. That is actuallymore precise than the precision of the suspendedsuspension bridge. It is nowhere; it is really brilliant atnothing. You might call it a meaningless certainty.

MALKAWI: There is also the question of computation.Does it help the process or not?

KOLAREVIC: David Leatherbarrow says it is adistraction.

LEATHERBARROW: No, just the reverse. But do nottry to do everything with just one instrument. Do nottry to look through one pair of glasses to see the wholeworld. The computer is brilliant at what it does, but tomake a beautiful building out of a computer, forget it.Just forget it. There are many other things you need,and it will not help you with those things. It is like Le

Corbusier’s Modulor — it helps you with some things, butfor the really difficult things it is useless.

MALKAWI: Let us try to think about the development ofcomputation within the past twenty years. In addition, let ustake a look at buildings in relation to their performance. Isthere a link between the development in computation and thecurrent “high performance” buildings that we see today?

RAMAN: I see that the intrinsic process has not changed.What computation has done is that it has had a liberatinginfluence on design. It is not that the computation takes overdesign or replaces it in some kind of way. What it does is itallows you to explore more — more features, morematerials, more situations with a greater degree ofconfidence than one might not have been able to do without.But it has not intrinsically changed the process.

WHALLEY: One area the computer has impacted on is theidea of three-dimensional form in design. It would have beenvery difficult twenty years ago to do some of the structuresand forms and shapes that we do now. You could have donethem, but probably not as efficiently. The computer hasslightly expanded the repertoire, or allowed certain areas tobe explored, which before were just not open. So it is a tool.It has broadened what we can do, but at the end of the day itis a tool.

KOLAREVIC: Let us then address the liberating dimensionof the technology that several speakers have referred to. AsBill Braham mentioned, the air conditioning in the 1950swas thought of as being a liberating technology. If we acceptthat what the digital now offers us is another liberatingtechnology, I wonder if we will be regretting these newliberties at some point down the road; maybe not. I think weneed to ponder these questions as we embrace thetechnological advances with enthusiasm.

ROBERT AISH (from the audience): I take up the point thatAndrew Whalley made about the impact of computers in thisprocess, in particular in the geometric quality. The nextstage, which I think might be more interesting to discuss, is

Page 241: B. Kolarevic-Performative Architecture

234

that, essentially, in creating these computer tools, weare looking to generate a representation that can beshared by both the creative architects who are dealingwith subjective aspects of the performance and theengineers who supposedly deal with more measurableaspects. If we can create tools that allowcommunication between the creative side and theengineering side, I think that will hopefully have a verypositive impact on this whole area.

MALKAWI: My argument is that computation withinthe past twenty years has tried to bridge the gapbetween the artistic and the scientific. Computation isslowly integrating those two or at least trying to bringthem together by enhancing communication betweenthose two extremes.

AISH: One really needs to go further than that. AsAndrew Whalley mentioned, the scientist is doing anexperiment over and over again, in comparison to theartist, architect or designer, who is inventing everytime. The role of science or engineering is not just torepeat the experiment, but surely to produce somecausal model, some explanation. If we can build thiskind of building engineering causal model into ourdesign tools, then these can be directly accessed andinform the design process. We can actually build theprediction of the experience into the design. That iswhat completes the loop because essentially we want todesign with feedback.

JEAN-FRANCOIS BLASSEL (from the audience): AliMalkawi has asked if computation had brought aboutsome changes. There were no computers in architectureschools twenty-five years ago, and now there is onlycomputers. That is not just a leap in the quantity; wehave a leap in the type of computation that one can doabout the physical behavior of a building. What one cando now is, in fact, to build a simulacrum, a simulationof a building, and experiment with at least someaspects of it. Whereas previously, the main thing onecould do is analyze it and then size it, give thicknesses,

give types of materials, and so on. Today that is quitedifferent. You can take a bunch of sticks or a strangeshape and you can try and decide what is going to behaveas a structure, which is a very, very different positionbecause of simulations of a different sample. This hasreally changed the way one could look at technical issues,and at the same time it opened up new possibilities forarchitecture and other fields, as we could now buildstrange shapes. We also see the changes in the way onecan explore the thermal behavior of unusual spaces.

All the fields were very good at basically perceivingonly one solution. You could only use a solution that isalready known and adopt it. But now we can play withsomething and see if it will work. There is a “black box”danger in this, but clearly something very new ishappening. We are trying to find out what to do with thenew computational technologies. That is the crux, ourreal goal.

CRAIG SCHWITTER (from the audience): Whileeverybody has a computer today and there is atremendous computing reservoir at our fingertips, themodels are still incredibly crude — we are in version 1.0.We can model extremely complex kinds of structuresurfaces, have all kinds of cutting patterns, and yet whatwe have actually seen is a very simple, elastic relationshipwith material properties. Things are still very crude.

Somebody mentioned that the juxtaposition of thequality and quantity is interesting. It is quitestereotypical of architects and engineers who use it.

Some of these more complicated geometries aresometimes actually driving deeper wedges betweenarchitecture and engineering. We often do not have theability of doing a feedback loop on a project. Sometimesarchitects do not even want to have feedback. Sometimesthese performative techniques actually lead to extremedivisiveness. It is nice to say that we all work with thesame model, but those of us who worked onmultidisciplinary projects know that we all parse off intodoing our own little models. It is rare that fabricatorsactually use the model that we give them. They might useit as a basis for their explorations, for developing their

Page 242: B. Kolarevic-Performative Architecture

235

own models that they can trust. So, sometimes thesedigital technologies can actually be quite divisive. Aswe look for things that can draw us together asarchitects and engineers, we need to be veryconscious of that.

MALKAWI: It can be dangerous to a certain degreeif one uses the computational tools as a black boxwithout really understanding the way it works.

SCHWITTER: I do not think one uses the finiteelement analysis when one is designing a structure. Ido not necessarily prescribe to the black box theory.Sometimes I think it is actually acceptable to usecomputation as a black box as long as oneunderstands the greater discipline that one is workingwithin, in terms of being able to ask whether it is theright solution or the wrong solution.

RAMAN: I would qualify that slightly by saying thatthe black box sometimes takes the place for an innateunderstanding of what is actually going on. Anexcessive reliance on the computer sometimesdetracts from developing an intrinsic field (or perhapssomething worse). What you are talking about is thatyou are not doing a finite element analysis becausethrough your experience you have a feel as to how thestructure works and what is likely to work and whatis not. I have sensed that for the new generation ofengineers coming through that understanding is beingreplaced by the reliance on this process.

LEATHERBARROW: My only point about thecomputer would be not to mistake the “menu” for the“meal.” I do not think the representation is the sameas the reality. That is the only issue; that is partlywhat Ali Malkawi is struggling with — the degree towhich the feedback from outside the system isadequately represented within the system. Now wecan build up and make increasingly moresophisticated representations and simulations, but atwhat point do we just forget the meal and eat the

menu? The representation, the simulation, are absolutelynecessary, powerful, effective, unprecedented — all ofthat is true. But it is still partial and there has got to besomehow, within the system, a recognition of its relativeautonomy.

The question of engagement seems to me the mostpressing one for performative architecture. Thesimulation is adequate under certain circumstances butwhen those circumstances change, the whole frameworkneeds to adjust or modify itself. It is the question oflimits — how much of this so-called feedback can beabsorbed into the instrumentality.

I do not think there should be any artificial limitsimposed on the research. In fact, there should be moreand more simulations. But I think there should also be asense that what has not been simulated will actuallyenlighten the project in the performative sense one isafter. But maybe we differ on this.

RAHIM: With regards to simulation and developing andevolving projects, it is an emergence between theenvironment and the user. That is in a certain sensesimilar to knowing what you are working with. Forexample, if you are working with a hammer, you becomea hammer, and if you are working with a temporal tool,you have to become one with it — and that is exactlywhat non-linear software allows us to do. It is to provideand provoke many conditions that you could not foreseewithout actually engaging the development of theprocess.

LEATHERBARROW: What about the representations ofthe conditions you could not foresee? Mine is only aquestion about the adequacy of the representation to thegiven condition.

RAHIM: That is the question. If you think of thedifference between the virtual and the real, that is wherethere is some very interesting potential because you haveto go through the process of actualizing something withvirtuality. The process of actualization is a mediationprocess with which we operate in a given moment in time.

Page 243: B. Kolarevic-Performative Architecture

236

The representation here itself is not limited becauseone can run a hundred scenarios and the formationproduced contains these virtualities which are multipleand not singular. The goal is not to reduce thesevirtualities into the singular, but to maintain amultiplicity in the formation.

ANDRE CHASZAR (from the audience): There is arelationship between simulation and exploration. Sowhen one talks about performative designing, often itseems that the desire is to arrive at some sort ofemergence. That something would come out of it thatone had not foreseen. Is that valid from the point ofview of engineering analysis? If I understand MahadevRaman’s point of view, you should know the answerbefore you begin, and so when you run your simulationand you get a result that you did not foresee, then yourfirst thought would be that the simulation is doneincorrectly.

RAMAN: I had exactly that experience within the firsttwo years of practice where I was the new kid whoknew about computers and punch cards and thosethings. So having gone through an elaboratesimulation for one of the Richard Rogers’ buildings, Iproudly presented my cooling load calculation freshlyoff the computer to a crusty old engineer who said Iwas out by a factor of four. Sure enough, I had madesome mistakes and we eventually got to the pointwhere I got it right, both to his satisfaction and tomine because I did actually discover the mistakes. Butthat it is not entirely clear-cut because sometimes thesolution emerges through the exploration of differentscenarios and the optimization process. There is alsothe ability to recognize that solution as it emerges,which is something that is outside of the simulationand is innate in the experience of the practitioner. Theway in which a problem is presented to you inengineering school is: if this, this, this and this, what isthe answer? You use a little bit of experience and dothe calculation, you come out with the answer and theprofessor can tell you whether it is right or wrong.

When you go to the design office, working with a group ofengineers and architects, you find that what you do notknow outnumbers what you can define. So in theory youare right. If you could define all of the parameters that gointo a process, and if you have a simulation of the universethat is correctly dealt with, you could set that off andeventually come up with something that says this is theanswer. In reality you can do that, but you can do that onlywithin very limited areas where the boundary conditionsare definable within limits. Frankly, I do not think we willget to a time, in my generation, where you can model all ofthe complex issues that go into design in such a way thatthe computer can come up with the optimal answer.

KLOFT: You need a lot of experience dealing with theseprograms and these tools. On the other hand, architectsand engineers are coming closer together with these tools,as we are experiencing in our office, where we have botharchitects and engineers.

Engineers should not only be educated that this is rightand this is wrong. We normally cut sections through thebuilding and we do not think like the architects in a spatialsystem, which is especially true for “freeforms.” In ouroffice, much of the work with freeforms is done byarchitects. So, there is a chance that the tools can bringboth closer together in the future. But I do not think thatone tool can bring the whole solution.

BLASSEL: Some of the models can cast more light onsome aspects at the beginning of the design. Perhapsarchitecture is not really so much about the space as it isabout judgment, about the synthesis. It is about taking awhole range of questions, weighing different things. Wherethese computations become interesting and so confusing isthe fact that you have to bring in all these things anddecide which is more important — the energy efficiency inthe building or views from it, etc. There are just so manythings to consider. The computer models may help usdecide. They may help us to understand better.

Page 244: B. Kolarevic-Performative Architecture

237

18OPERATIVEPERFORMATIVITY

FRIED AUGENBROEJEAN-FRANCOIS BLASSELJAN EDLERPETER McCLEARYGREG OTTOLARS SPUYBROEKALI M. MALKAWIBRANKO KOLAREVIC

Page 245: B. Kolarevic-Performative Architecture

238

18OPERATIVEPERFORMATIVITY

FRIED AUGENBROEJEAN-FRANCOIS BLASSELJAN EDLERPETER McCLEARYGREG OTTOLARS SPUYBROEKALI M. MALKAWIBRANKO KOLAREVIC

Page 246: B. Kolarevic-Performative Architecture

239

MALKAWI: I want to reflect on the original idea ofhow we came to think about this topic and to brieflyintroduce the subject that we will discuss in this panel.

Performance in architecture is not new and is avery general concept that has been discussed indifferent forms. From a theoretical perspective, theconcept of performance has been discussed in thissymposium from phenomenological and structuralperspectives. Our invited speakers come from a varietyof disciplines and we tried, as much as possible, toselect individuals who would bring various views fromthe conceptual and operative sides of performance. Inhis presentation, David Leatherbarrow provided acomprehensive introductory framework forperformance from a theoretical perspective. Theduality of the subject was enforced by Peter McClearywhen he discussed issues of negotiation between theopaque and the transparent, and between art andscience. He provided a theoretical framework forreflection on the complex nature of architecture. Wehave witnessed this negotiation or mediation betweencomponents being discussed in a variety of formsduring the symposium. One of the main issuesdiscussed as a possible factor in bringing new forms ofarchitecture is the computational instrument which Ibelieve can act as a mediator or an interface betweenthese two worlds, art and science.

Fried Augenbroe talked about the issue ofmeasuring performance. This is a topic that we havebeen working on for a long time. The topic has a longhistory from theoretical as well as practicalperspectives. One of its byproducts is thecomputational instrument that has been used in designand analysis. Although there has been majoradvancement in this area and we see its influence onthe profession, there has been a “disconnect” betweencommercial tool developers and the designers.

We see two types of users, the architects as wellas the engineers or the consultants. We do have toolsto answer specific questions from an analysisperspective, but we lack the tools that aid in synthesis.Our tools also lack the understanding of the process

that supports the design analysis as buildings are designedand constructed.

We discussed performance from an aesthetic andcommunicative point of view, such as the façade project.Discussions also included the cultural importance ofperformance in regard to buildings. The Buro Happoldgroup talked about tools and their adequacy andvalidation. Do performance numbers provide an absolutemeasure which is always correct? Different problemsrequire different solutions and, in most cases, measuringperformance can be comparative; relative performance iswhat is needed in many instances.

Lars Spuybroek described the measurements of motionin relation to space; Jean-François Blassel addressed thecoupling of different types of software to answer questionsof performance. Finally, Peter McCleary provided aperspective on how the different actors of the design of abuilding interact although they have different languages.

We attempted to separate the discussion in the panelsfrom the operative as well as the conceptual perspectives,although we understand their interactions. The operativecomponent, which is the topic of this discussion, is relatedprimarily to building operation and its tangible measure ofperformance, a performance that can be judged by relativebenchmarks. Operative performance, as we see it, isclosely linked to the computation instrument and this iswhat I would like our panel to focus on discussing.

The first question for the panel is related to theinstruments or the analysis tools and how they influencedesign. What is the role of the analysis tools, as it standsnow, in influencing design performance? Try to reflect on itfrom your own domain of expertise.

AUGENBROE: I have no direct design experience myself,but I can reflect a little bit on your question. You and I,and many others, have been working for many years onimproving building simulation tools and we have neverreached a more mature stage than we have at the moment.At the last international building simulation conference,where the latest tools including CFD (computational fluiddynamics) were presented, you could conclude there wasvery little in terms of functionality that was still to be

Page 247: B. Kolarevic-Performative Architecture

240

desired. And I remember, in the closing session, one ofthe overriding conclusions that we came away fromthat conference with was that simulation is so mature itcan eventually become invisible.

So, that leads to the next point and that is howdoes simulation then support design? I think, at themoment, the only answer is: in a very limited fashion.And why is that? I think because of the lack informality. We use simulation tools to study behaviorduring design evolution, but we do that in a way that isnot well coordinated. My conclusion, as one of thepeople who wants to look beyond what we do withcurrent matured simulation, is that we need to developsomething which puts simulation and analysis in amuch more formal framework. What I suggested in mypresentation was to base this framework on predefinedvirtual experiments. We need to define clearexpectations of the design and be able to measureproposed solutions against those. For this we wouldneed very matured simulation tools, which have to bepreconfigured in a way that you can perform thesevirtual experiments. I have not seen any platformsbeing developed to do that in an effective way, but Iknow many efforts in research that are trying to do justthat, leading to what you could call the next generationof simulation.

BLASSEL: The tools that we have now are aboutmaking abstract presentations of some aspects of thebuilding. But this is not a new situation. Currently wehave specialized tools that are part of a toolbox. We doapply them to specific problems. As Fried Augenbroesuggested, if there are similarities between projects,maybe we can call for an array of tools that can beavailable and conform to a benchmark which allow usto compare solutions.

But for the generative part of the project, I wouldthink that we have to use tools in a very discerningfashion to develop knives with which we cut theproblem, just as all architectural problems are verycomplex, with all the issues that are interrelated. Thesetools allow us to make a very minute dissection of what

we are trying to understand. Depending on the shape of theproblem, we need to cut in different places using differenttools, so we need to find specific tools to help us imaginesolutions.

KOLAREVIC: I would argue that the current simulationtools are completely useless from a design perspective. Ifyou do agree with that as a position, then you ask what willmake them useful. What could make them useful is perhapschanging the resolution. High-resolution tools demand anincredible degree of detail, if you want to get a usefulsimulation from an engineering point of view. We needsome kind of low-resolution simulation tools that areapproachable by architects.

The conceptual design is not about detailed modelingof the rooms so that you can do the CFD simulation. If weneed low resolution, then we would somehow need aninterface that does not demand such a high degree ofdomain knowledge. You almost have to be an expert inenvironmental design before you could use any of the CFDsoftware. What we may want to see from a designperspective is something that has a genuine dimension fromthe performance point of view. So, coming from a designperspective, what will make these tools useful to us? Rightnow I do not find the analysis software of much use inconceptual design.

BLASSEL: Well, maybe what you are trying to say is thatthe low-resolution tools require more abstraction — it islike a sketch before we do a design. What we need is theequivalent of a sketch, a technical sketch.

MALKAWI: This might be related to the larger question ofthe user of the tool. Is the user an expert or non-expert? Isit the engineer or is it the architect? Much debate has beenspent targeting this question during the past decade andthe consensus is that the simplified tools did not delivertheir promise of helping the designer. Is it more productiveto understand the process of design, including thecollaboration and communication between the differentparticipants, and develop tools that support these activitiesbetter? Identify the users within the process and

Page 248: B. Kolarevic-Performative Architecture

241

accordingly develop tools that satisfy the question ofresolution that support these users but do not focusonly on the architects.

SPUYBROEK: Every project starts with analysis andwith the analytical and the weighing of data; basically,you do that with every project. At the end, the projectalso ends with things that are weighed and thatactually cost money. But, in between, there is muchnon-metric stuff going on. I think the whole art isactually to move from the real that is existence, toanother real that is the real plus your project. But youhave to pass through something that is not real or thatis actually abstract. Of course, that is the wholequestion of the diagram and the different shapes inbetween organization and structure.

So, we would have the world of organizations ordiagrams and you would have the world of structuresthat actually can be quantified. But there is thispassage, it has to pass through a stage of order andqualification on an abstract level. The question is howdo you go from tools that are only capable of cuttingthings up to a tool that can actually synthesize andintegrate and reach an abstract level.

MALKAWI: On the note of abstraction, I think itmight be worthwhile to ask Peter McCleary about histhoughts.

McCLEARY: Aristotle said something similar on thesubject of skill. He believed that one should becomesufficiently informed to understand the playing of themusician, the rhetoric of the poet and the “how” of thebuilder. If one becomes too skilled, it is not possible todifferentiate between the master and the slave, and itwas never our intention to be a slave to the mediation.

Recently, after my short talk on the relationshipbetween structure and space, one of my distinguishedcolleagues said that there was no longer such arelationship. In this era of Frank Gehry and otherBaroque architects, the building’s internal surfaceshapes, or is shaped by, the activity space, and the

external surface derives its shape or configuration fromthe geometry and scale of the city context. The structure issandwiched between and ranks lower in importance thanthe spatial relationships for the inside and the outside.

There is a proportional relationship among the threedimensions of the volume of the space framed by astructure. This does not mean that the height is theharmonic mean of the two dimensions of the plan, as itwas for Palladio. In plan, we weave the warp and the weftof both the space and the structure; their proportionalrelationship can be described in the language ofmathematics in general and geometry in particular. Justas we can speak of the geometry of the space and of thestructure, we can describe the geometry of sound and thegeometry of light. There is much work to be done on thesynthesis of those geometries. In the absence ofoperational techniques, it will remain the task of maturedesigners to negotiate and coordinate that synthesis.

KOLAREVIC: That ties back to the issues ofrepresentation that were touched upon several times. Theissue of representational integration relates to the issuesof geometry that you are bringing up. That is anothersignificant missing element in the visual tools that we haveaccess to. Perhaps these different geometries that PeterMcCleary is referring to could somehow be broughttogether in this visual domain. The reason I bring these upis that I was surprised by Fried Augenbroe’s assertion thatwe are done with building simulation. His suggestion iswhat we have is already sufficient and mature. I thinkthere is quite a bit to be done in that realm.

MALKAWI: Knowing the work that is being done in thefield of building simulation, which is not just limited todeveloping tools for the architects, I would like to giveFried Augenbroe an opportunity to respond and talk aboutthe representational aspects of the simulation and thework that he has done on interoperability.

AUGENBROE: I would like to respond to your earlierstatement that tools are completely useless. No, I do notagree with what you said. I think you were close to saying

Page 249: B. Kolarevic-Performative Architecture

242

that tools are completely useless to designers, but yousaid they are completely useless to design.

KOLAREVIC: That is correct.

AUGENBROE: Which one did you mean?

KOLAREVIC: Both.

AUGENBROE: I think today we have many tools thatare certainly not useless to design. Whether they areuseless to designers is a loaded question. There is anold discussion stating that we are placing the wrongtools in the wrong hands or we are making the wrongtools for the wrong purpose. I think we are actuallystill doing that.

One of the things that is haunting us in buildingsimulation research is that we are developingfunctionally undefined tools. We are so focused onstudying the physical behavior that we are first andforemost trying to capture that behavior in our tools,and then count on the expert and whatever divineinspiration occurs in a given project to use the tooleffectively. Reality has it that tools are usedregardless of whether a certain simulation actuallymakes sense. And that is where Branko Kolarevic’sstatement comes from, is it not? If so, I agree withhim, in a sense, but at the same time I am amazed tohear designers requesting tools that work inconceptual stage and maybe later on too. This is incontrast with a movement I see going on in corporateAmerica where more and more firms are delegatingrather than incorporating the expert engineeringanalysis work. So, why would design firms try to goagainst that trend — even if the tool is very simple? Ithink that designers will stop asking for simulationtools once they realize that if you have a trustedmedium under the next mouse click, this medium willgive you more accurate information about thebehavior that you are trying to study to make the nextstep in design, and more quickly than you could everhave generated yourself.

MALKAWI: I would like to give Greg Otto one last wordbefore we move further.

OTTO: I would like to try to link together the interestingconversations we have had so far. I think design isactually changing and I think we have changed into amore collaborative state. I am finding that in manyprojects that I get involved with now, the architect has anidea, it is only an idea and now he wants to sit at the tableand he wants to sort out the design. In reference to theCFD model or the simplified version, I do not know if thatis exactly what you want. Do you want a machine to sitthere or do you want somebody sitting at your table with aworld of experience, informing you with what are themerits or what are the downsides of this, that or theother? So, with regards to tools, I think that the currenttools that we have allow us to look at more scenarios andthey are going to only get better. But, you will always needthe people to interpret the results and that is the roundtable effect. I also believe that this is going to lead to aperformative architecture because it is influenced by aworld of knowledge, not a singularity of knowledge.

MALKAWI: At this point I would like to invite questionsfrom the audience.

UNKNOWN (from the audience): I am not sure if we losttrack of that idea of the usefulness of lower resolutionanalysis or simulation tools. I hope that we have not. Ithink Greg Otto correctly pointed out that you want to dothis in a collaborative environment. So, if you aredeveloping a lower resolution analysis tool, the purpose ofdoing so is not to enable the architect to do the engineer’swork. But it would actually help the engineer fill thatcollaborative role earlier in the process. So, conceptualdesign in the absence of those tools still happens in thebrain and on the paper. It is only late in the project, oftentoo late, that you can leverage the power of theseanalytical tools. You can make them simplified to thepoint where they can be used without knowing too muchalready, but still getting useful answers. Then you can geta better use for the tools.

Page 250: B. Kolarevic-Performative Architecture

243

MALKAWI: We have two issues on hand: the tools thatare related to the process and probably different types ofresolution and the interoperability between the toolsthemselves that will support the process itself. This hasan effect on the issue of accuracy required by the tools. Iwould like to call on the panel to get their thoughts inregard to these issues. I would like to start with FriedAugenbroe as he and I have been working on these issuesin our research.

AUGENBROE: It is not for lack of trying that we do nothave tools that do something meaningful in earlyconceptual design stages. A few years ago we did aninteresting project to find out whether current tools areaccurate enough if you look at them from an uncertaintyperspective. We found that there is a technique calledprobabilistic inversion where you try to find out what isthe simplest tool that would still give you the sameaccuracy in the results given the fact that we work in anuncertain world with uncertain input. We then provedthat most of the tools could be much simpler than theones we are working with. So, given the fact that yourinput is very uncertain, a simpler tool may give you thesame level of resolution as the more accurate tools. Thisis an interesting and somewhat uncomfortable conclusionfor tool developers. Now, how does that translate tomaking tools simpler or having a lower resolution(which, by the way, are two different subjects)?

The 1970s were dominated by attempts to developsimplified tools — dumbing them down, so to speak, sothey could be used by people who were not experts. Ithink that movement has virtually died down. A moreinteresting area that we are working on now are toolsthat still reflect the basic physics of the field and are notnecessarily that easy to use, but require much lessdetailed input. Now the problem is that if you look at theuncertainty levels in the very early design stages withregard to information availability, you will find that youhardly get any resolution from these tools to supportdesign decisions. If this is ignored, they become verydangerous because they deliver some kind of quasi-certainty.

I think that we should do more work on definingsimplified experiments that can be applied in the earlyphases of the design. They should require only verylimited input, so you have to carefully design thoseexperiments, and then make simulation tools to executethem. That, in my view, is the way to go, not taking thesimulation tools that you already have and trying tomake them so simple by defaulting and doing other kindsof things to them to make them usable in a very earlystage when you have much less refined knowledge of theinput data. I think that will not work. So for the firsttime we are in the stage where we can define the kind ofperformance quantifications that we are looking for. It isreally back to the drawing board. But the point is that allthe projects that we saw debated at this symposium areso far out of the routine that it will be an illusion to thinkthat the development that I described will give you anyuseful support for those kinds of projects in the nearfuture.

SPUYBROEK: We are discussing tools as if they aretowards determinism or to a purpose. Marx, of course,already knew that a tool was not for the user but actuallyproducing a user. That is the first product of a tool — itsuser. It is not that we were users and then all thesoftware companies are designing software for us. It isnot true. It is actually the other way around, and that iswhy I want to argue a bit against instrumentalism andargue against this whole idea of mediation where there isA, there is us, and then there is the world that is B. Wejust mediate and bridge phenomenologically with our toolthat is neutrally hanging in-between. I do not think thatis true, even for a hammer; something so overly definedby its function to hit a nail in the wall is every now andthen used to kill somebody. So that is already goingbeyond the profile of the thing.

MALKAWI: I think we are not articulating our needs asdesigners to the industry or the tools’ developers. This isanother discussion; however, I would like to ask PeterMcCleary to respond to Lars Spuybroek regarding theissue of mediation.

Page 251: B. Kolarevic-Performative Architecture

244

McCLEARY: In this context, there are at least twointerpretations of the phenomenology of tools. Whilehumans and their equipment are part of a contextualtotality, tools are either equifinal, that is different toolscan serve the same end or purpose, or are equipotential,that is the same tool can serve different purposes. Thus,a hammer does not insist that the world is a nail; or, asthe Chinese say, it is the one that sticks up that getshammered on the head.

KOLAREVIC: What are the kinds of potentials andopportunities, and what are the pitfalls that we shouldtry to avoid when thinking about the performative inarchitecture?

SPUYBROEK: Architects never see themselves asspecialists, and that is basically part of the problembecause they are normally at the table as generalistswith specialists helping them. The specialists aredealing with the hard data and the architect alwaysjumps on the table and puts it all together, but nobodyreally understands how. That is difficult. It is not abouthaving the engineers conceptualize and going up a levelhigher, so they can actually conceptualize with usduring the early stages of the design. It is actually theother way around.

AUGENBROE: The last two questions are closelyrelated. What we are really probing is the way you seethe whole building process being more managed andwhat is the role of the architect in that, and what kindof collaboration environments will ultimately work as acatalyst to change that process. The interesting part fordesigners is how to anticipate a future where, indeed,collaboration is the driver of the process. You get intoissues of formality, predictability and coordination,based on integrated representations, linked tointegrated performance taxonomies. We have toprepare our profession for this future, if indeed that isthe future, of very tight and coordinated integrationwith all of the other involved disciplines. If the architectwants to play the role of the design manager — as I

think he or she should and might want to — the architecthas to come to the table with the tools and representationsthat allow him or her to manage the performance contracts,making explicit in what way the engineers have to supplythe right expertise. I think these two questions point to adirection where I think performative architecture should go;keep a close watch on what is happening in the industry interms of collaboration contracts and project deliverysystems.

UNKNOWN (from the audience): I would like to knowwhether we are asking too much of the architect or askingtoo little in how much we are requiring him or her to dowith regard to deciding how a project moves forward. Forhundreds of years, perhaps thousands of years, we were ableto design buildings that met their performance objectives.We were able to do things reasonably well, and perhaps wewere able to do that because of the types of materials weused. We did not necessarily design things that would notwork. It seems now that we are able to design things thatwould not work and we have to hand it to an engineer whois going to undo the problems that we have created. Forexample, it seems to be ridiculous at times that we designglass boxes and put them in the desert and ask ourengineers to find a way to make them work. In some waysare we doing such things because we can do things that wecould never do before? Are architects now only generalistsand so are not contributing much to the profession?

McCLEARY: I have been lucky to work with somearchitects who were both generalists and specialists. At themeeting of design collaborators, the architect had thebroadest and deepest culture. Not only were they generalistsbut they also had particular expertise. At that meeting itdid not make sense to focus on or even talk of thepercentage of lime, silica and alumina in the cement. Rarelydid we reach that level of detail. When we did, no otherperson understood its significance. All participants wereasked to bring something that could be synthesized into awhole. Since others would not understand, the architectmight not talk about mood and feeling. However, they coulddefine the characteristics of the activity and what kinds of

Page 252: B. Kolarevic-Performative Architecture

245

spaces might accommodate that activity; even a spacethat “inspires an activity.” While he or she might notknow the measure of the space, he or she had a goodsense of the orientations or isotropy of the space. Inaddition to space planning for activities, the architectmight propose a new structural configuration, or evena new way of living in space. Thus, the architectbrought a considerable degree of oversight expertise tothe collaboration. Their proposal was not a building,represented graphically, but concerned those thingsthat have to do with habitation. No one else did that.So, yes, the architects were generalists but they alsobrought a lot of particular skill and knowledge to thetable.

AUGENBROE: Getting here, I noticed that the mainfocus of the designers was to bring their design to thepublic stage and have it perform. To me, that wassignificant. This showed a kind of bias in thechoreography, with focus on seeing the object performon the public stage rather than regard performativityas the capstone for the matching of client expectationsand fulfillments. From my perspective, I think thatmanaging design choreography is ultimately moreimportant from a research and dialogue point of viewthan having the building as an object perform on apublic stage. But that may be strongly biased by myengineering background.

BLASSEL: I would like to go back to the question ofthe glass box in the desert, but not so much to providean answer about the glass box but about what is goingto follow. That is, why can’t we just do it the way it hasbeen done before? Maybe it is because we findourselves in a global position and, at the same time,find ourselves in a comfortable situation which we arenow enjoying in the United States or in WesternEurope. But we do not have enough resources;something needs to be done in a much more inventiveway using the resources more cleverly, using energymore cleverly. I am not saying that technology will bethe only way to address this problem, but it is one of

the ways in which it can address the problem. So, thetools, and the fact that we have to try to use them andtry to make them evolve and understand the way inwhich the physical nature of the building interacts withits more spiritual nature, is essential.

EDLER: One of the main issues Peter McClearydiscussed, and that I agree with, is communication. Thisis an aspect that we did not discuss much and it isblocking many powerful materialities of architectureright now from appearing. It is a very pragmatic issue.As I started as an inexperienced professional — atypical architect coming from university at some pointhitting my first real construction site — what Iexperienced was that energy gets lost in architecture atthe point when you try to erect it. Suddenly, you have asituation which is purely about competition and beingresponsible for something. Who is the person to takeresponsibility? This is a very crucial issue because, as Iremember on the construction site in Graz, the maintopic was basically trying to take as little responsibilityas you can because it is all about money and this iscausing so much friction. If there would be a possibilityof getting this friction away from the production ofarchitecture, I think architecture would be performingmuch better.

McCLEARY: Even though everyday the performanceaspect of each technique becomes more resolved, I amnot persuaded that everything is measurable. I am notpersuaded that things that have not been measured arecapable of being measured. One of the architect’sresponsibilities is to reveal aspects of our environmentthat seem to be hidden or concealed. For example, watercan be fixed in place or I can carry it around with me —to drink or it can be conceived as H2O. It also has ameniscus, it has depth, it has pressure, it can flow, andso forth. There are many ways to interpret an object.One of the major responsibilities of the architect is toreveal, through designing and building, the nature ofhumans and their worlds and the dialectical relationshipbetween them.

Page 253: B. Kolarevic-Performative Architecture

246

SPUYBROEK: Today, performance is related to formand information. I see performance as something thatcomes after function and after the event. We hadyears of function where bodies were mechanical andmechanistic. You had necessity, you had necessarybehavior, mechanistic behavior, and then you had play.You had this space of accidents where you had playand multifunction. I think performance is a real issuein the sense that it is trying to look for actuallymerging these. Because we lost some time duringpostmodernism and deconstruction where architecturebecame a language, now there is a real interest. Allthe conferences are about electronics or about thebody, and there is a real sense of materialism. It is nota reductionist materialism but it is looking at the realmateriality of experience, feelings, being andstructure. What does architecture actually do? Whatis the operationality of it, instead of its aesthetics andits language? I think this is performance but Iunderstand there is something in between which ismono and multi, something that is not eitherdeterministic or totally wild but is actually somevague, wider sense of redundancy. Not free play orpure function, but is actually a widening up offunction that makes it relate back to bodies and tostructure.

OTTO: Peter McCleary’s remarks provide for a verygood closure. The world that we face is growing morecomplex and I think in that complexity we are forcedto deal with more issues and, as Peter noted, we canmeasure the scientific but there are many other thingsthat are not scientific and cannot be measured. To me,performative architecture is trying to find ways andmethods to understand and deal with those issues in away that, ultimately, ends up with a result that issatisfactory to the design intent.

MALKAWI: In my opening remarks for thesymposium, I stated that we tried to organize it notonly with the hope of raising questions regardingperformance, but with the attempt to find common

threads between the different views presented regardingperformance. I think we definitely raised the questions andwe found common threads at least conceptually. We didnot and we are not trying to synthesize all the differentaspects and views of performance as being projected fromour discussions. As many of you stated, both DavidLeatherbarrow’s and Peter McCleary’s presentationsprovided the needed anchors to introduce and close ourevent in regard to performance in relation to architecture.What I would like to do now is give my colleague BrankoKolarevic the opportunity to provide his final remarks.

KOLAREVIC: I would just like to remind you where westarted. Ali Malkawi and I have positioned this symposiumas a fantastic territory that lies in what we describe as twopoles — the pole of the conceptual, or what we call theperformativity of the virtual, and the pole of the operative,or what we call the performativity of the real. I think themetaphor of the poles is an appropriate one because itrelates to the sphere. And a sphere, as an endless entity,can have more than two poles, depending on how youqualify it. I think each of you were able to define your ownpoles, oppositional poles, in the discourse that we had overthe past two days. We were not hoping to actually getsome answers out of this event. As Ali Malkawi stated, ourintent was to actually generate more questions than theanswers. In that respect, I think, we have succeeded.

Page 254: B. Kolarevic-Performative Architecture

247

ALI M. MALKAWI

EPILOGUE

Page 255: B. Kolarevic-Performative Architecture

248

ALI M. MALKAWI

EPILOGUE

Page 256: B. Kolarevic-Performative Architecture

249

This book presents varying views on the concept ofperformance as it relates to buildings. Engineering andarchitectural perspectives, theoretical interpretations,as well as ideas from research and development, wereprovided by scholars, researchers and practitioners.These discussions suggest that “performance” isunderstood differently by the different participants inthe design of buildings.

A driving force behind this assemblage was thedesire to investigate the impact of the recent surge incomputational instrument development and use inarchitectural practice and “high performance”buildings. Although the performance concept is not new,and is inherently a basic component in building design,its application to the building envelope and structurethat is represented in “high performance” buildings isrelatively recent.

The book illustrates that buildings embodying thenotion of “integration” or “high performance” conceptsare developed through highly interactive relationsbetween architects and engineers. This integration isinfluenced primarily by the “process” of collaboration.This process has been imbued by advancements incomputational technology that narrowed the gapbetween the design actors, including architects andengineers. It made it easier for engineers to be engagedin all aspects of design activities, and be able toinvestigate and clarify questions by conducting fast,robust analysis. This allowed new forms of architectureto be possible.

For buildings that utilize “integration” or“performance” as their main theme, the process ofhaving the designers orchestrate the design integrationprinciples, and having the instruments bridge the gapbetween the participants, has proven to be essential.

Currently, computational instruments are used for analysisto support design decisions. They are used to check solutionsand to convince clients. The architects typically integrateperformance principles early in the design process. Theseprinciples affect both the form and the structure of thebuilding, and they get to be refined and optimized by theengineer’s active participation within the process using suchinstruments. Issues that used to be difficult to predict arenow possible within a relatively short period of time.

Although the profession has access to sophisticatedanalysis tools, work is still needed in many areas to facilitatebetter integration between architects and engineers, as wellas among the engineers themselves. Building professionalshave not investigated the full potential of computation to aidin the opening of new horizons for building performance.Integration of the advancement in computational theory andbuilding simulation is still in its infancy. The problem iscomplicated by the fact that the development of these issuesis driven by “immediate” market needs rather than long-term goals. New commercial instruments are beingdeveloped in an attempt to respond to users’ needs. Theseinstruments still lag in their progress due to a “disconnect”between funded efforts and commercial tool developments.

The book illustrates that, although the instruments havebeen maturing, only elite practices are utilizing them. Thesepractices have shown that these instruments can be a majorinfluence on the way buildings are conceived, designed andconstructed. Although relevant chapters of the bookillustrate how technological developments in computationalanalysis make it possible to bridge the gap between theengineering and architectural realm, much more can beaccomplished with a focus on the development of tools thatwill provide design synthesis rather than only analysis. Themulti-criteria and complex nature of building performancepredictions illustrates the need for more work in this area.

Page 257: B. Kolarevic-Performative Architecture

250

Page 258: B. Kolarevic-Performative Architecture

251

PHOTO CREDITSINDEX

APPENDIX

BIOGRAPHIESAUTHOR

Page 259: B. Kolarevic-Performative Architecture

252

PHOTO CREDITSINDEX

APPENDIX

BIOGRAPHIESAUTHOR

Page 260: B. Kolarevic-Performative Architecture

253

BIOGRAPHIES

GODFRIED AUGENBROEAssociate ProfessorGeorgia Institute of TechnologyAtlanta, USA

Professor Augenbroe received an MSc cum laude in Civil Engineering fromTU Delft in the Netherlands in 1975. He has spent most of his academiccareer in Europe, where he pioneered a commercial finite element toolboxfor building simulation and managed large European Union-funded projectson engineering interoperability and semantic building product models. Hehas led international consortia of academic researchers and industrialdevelopers, such as the COMBINE effort (Computer Models for the BuildingIndustry in Europe), a project spanning 1990–95. This project has deliveredthe first prototypes of the next generation of integrated engineering designsystems with an emphasis on building services engineering.

Professor Augenbroe has been active in the pursuit of internationalcollaboration, exemplified in prolonged working stays abroad at CSTB,France (1988), Lawrence Berkeley Lab, USA (1989), and at UNPHU,Dominican Republic (1996). More recently he has held visitingprofessorships at Loughborough University, UK, and the University ofNewcastle upon Tyne, UK. Since 1997 he has been heading the BuildingTechnology track in the Doctoral Program in the College of Architecture atthe Georgia Institute of Technology in the USA.

Professor Augenbroe teaches graduate courses and conducts researchin the fields of building performance concepts and simulation, the control ofsmart systems, e-business, system monitoring and diagnostics. He has alsoestablished an active research record in web-hosted collaboration andknowledge management, dealing with the development of software tools andtheir business integration.

Professor Augenbroe is on the scientific board of five internationaljournals, is American co-editor of two other scientific journals and haspublished over a hundred refereed papers. He has chaired three majorconferences and delivered six keynote lectures at international conferences.He was the chair of the IBPSA BS2003 conference in August 2003, and heis currently the coordinator of the e-HUBs project (IST-2001-34031). Theproject focuses on the tactical decision-making that prepares e-engineeringpartnerships, which are becoming ever more vital for the effectiveglobalization of the building industry.

http://www.coa.gatech.edu/phd

JEAN-FRANÇOIS BLASSELDirectorRFR Consulting EngineersParis, France

For more than ten years, Jean-François Blassel has been oneof the directors of RFR, the unique architectural engineeringfirm founded by Peter Rice in Paris.

Jean-François Blassel trained as an architect and anengineer, and has collaborated on many large-scale structureswhere architecture and technology are intertwined from thebeginning of the project.

He has worked in Europe, North Africa, Asia and theUnited States on large-scale buildings and structures. Theseprojects include the Kansai International Airport with PeterRice and Renzo Piano, several high-speed train (TGV) stationsand viaducts for the French railroad authority, as well as manysmaller technologically-oriented projects.

Jean-François Blassel also teaches at the Ecoled’Architecture, de la Ville et des Territoires in Paris and hasbeen a visiting faculty for the “Emerging Technologies”program at the Graduate School of Fine Arts of the Universityof Pennsylvania since 1999.

Jean-François Blassel received the 1994 FrenchGovernment’s “Album de la Jeune Architecture” Award forarchitects under forty, and he writes regularly for the Frencharchitectural press.

http://www.rfr.fr

Page 261: B. Kolarevic-Performative Architecture

254

WILLIAM W. BRAHAMAssociate Professor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

William W. Braham is an Associate Professor of Architectureat the University of Pennsylvania. He received a degree in CivilEngineering from Princeton University and an MArch and PhDArch from the University of Pennsylvania, where he has taughtsince 1988. At Penn he teaches graduate courses on light andenvironmental technology, and he coordinates the second-yeardesign studios. He practices with Studio Luxe, an architecturaldesign and consulting practice, and he sits on the boards ofIvalo Lighting and of Praxis, Penn’s design practice unit. In2002 he published a book called Modern Color/ModernArchitecture: Amédée Ozenfant and the genealogy of color inmodern architecture (published by Ashgate). He is currently co-editing a collection called Rethinking Technology: A Reader inArchitectural Theory, due to be published in 2005, and he iswriting an architectural account of environmental conditioningsince 1968.

http://design.upenn.edu/~brahamw/

JAN EDLERDesigner and Architectrealities:unitedBerlin, Germany

Jan Edler studied architecture at the Technical University Aachenand at the Bartlett, University College London. He graduated in1997 as a Diploma Architect. Since 1996 he has worked as a co-founder for the Berlin-based art group [kunst und technik]. In 2000he founded the architectural design studio realities:united (realU)together with his brother Tim Edler.

The team deals comprehensively with the staging of culturalevents and in the designing of material and information spaces.realU does research for the development of new technologies andprogressive working methods, ideas, messages and communicationstrategies. realU employs a great variety of scientific, commercialand artistic methods and strategies — often mixing and blurringconventional discipline-related work strategies.

Since 1998 realU has been focusing on projects researchingthe integration of media technology in the “art space,” both for self-commissioned and commercial projects. Current projects includeexpertise and strategic design work for the integration of media andcommunication technology for the Kunsthaus Graz, Austria, the newPaul Klee Museum in Bern, Switzerland, and a research project onprototypical video communication for the Bauhaus-Dessau,Germany.

Their projects were presented at numerous internationalexhibitions and conferences, and have been published worldwide.They were awarded several internationally recognized designdistinctions.

Recent teaching positions held by the duo include theDepartment of Architecture at the Technical University of Berlin(2000–01), the Bauhaus Foundation, Dessau (2001–02), and thePasadena Art Centre College for Design, Los Angeles, USA (2003).

http://www.realities-united.de

Page 262: B. Kolarevic-Performative Architecture

255

THOMAS HERZOGPrincipalHerzog + PartnerMunich, Germany

Professor Herzog has been the Dean of the Faculty ofArchitecture at the Technical University in Munich (TUM),Germany, since 2000, and he has been a Guest Professor atTsinghua University in Beijing, China, since 2003. Hereceived his Diploma from TUM in 1965 and earned hisdoctorate in architecture at the University of Rome “LaSapienza” in 1972 on “Pneumatic Structures.”

Thomas Herzog founded his practice in 1971. Sincethen he has worked jointly with Verena Herzog-Loibl. Hiswork has been focusing on the development of buildingsystems for the use of renewable forms of energy, thedevelopment of new building products, housing,administration, industrial and exhibition buildings, etc. Hewas a partner with Michael Volz from 1983–89 and hasbeen a partner with Hanns Jörg Schrade since 1994.

Professor Herzog was the Guest Professor at EcolePolytechnique Féderal de Lausanne (EPFL) and recently theGraham Professor at the University of Pennsylvania. He is amember of many professional organizations, including theInternational Academy of Architecture of UNESCO, Sofia,Bulgaria, and the Academy of Sciences and Arts, St.Petersburg, Russia. He has won many honors and awards,including the Mies van der Rohe Prize in 1981, the UIAAugust Perret Award for Technology in Architecture in1996, the Grand médaille d’or d’architecture, Académied’Architecture, Paris, in 1998, and the European Prize for“Solares Bauen” in 2000. He is the author and editor ofmore than a dozen specialized books in seven languages,including seven monographs. He was the German GeneralCommissioner of the International Biennale of Architecturein Venice in 2000 and has been the Expert “DeutscheForschungsgemeinschaft” since 2000.

http://www.herzog-und-partner.de/

HARALD KLOFTPrincipalosd – office for structural designDarmstadt, Germany

As an engineer, Harald Kloft has a passion for the close relationship betweenstructural engineering and architecture. He promotes the idea of structuraldesign in which structural principles are integral to the architectural designprocess. He defines structural design as a process “supporting thearchitecture” in which structures are developed and materiality is researchedto convey the architecture’s ideas. Such understanding of the structuraldesign was the basis for his undertaking of the engineering leadership in theoffice of Bollinger + Grohmann for some challenging “non-standard”architectural projects, such as Bernhard Franken’s “Bubble” and“Dynaform” pavilions for BMW, and Peter Cook and Colin Fournier’sKunsthaus Graz, Austria.

In 2002, Harald Kloft founded the office for structural design (osd)together with Sigurdur Gunnarsson, Klaus Fäth and Viktor Wilhelm. osd isan engineering practice whose work is focused on the integration of innovativeaspects of structural design in architectural projects, such as experimentationwith new materials and the digital control of a design process that allowssimultaneous formulation and the examination of ideas. Current projectsinclude the Klöpp Ministry Buildings in Reykjavik, Iceland (architectBernhard Franken), a new RegioTram Railway Station within the centralstation in Kassel, Germany (architects Pahl/Weber-Pahl), and a new InstituteBuilding for the German Center of Aviation and Space Travel (DLR) inStuttgart, Germany (archtiects bk + Arno Brandlhuber).

Aside from his practice, Harald Kloft is continuously engaged inacademia. He received a doctoral degree at the University of Technology inDarmstadt, Germany. Since the mid-1990s, he has lectured in many countriesand has taken part in various research and teaching programs, transmittinghis experiences and knowledge of practice. He is a Visiting Professor at theStädelschule in Frankfurt, in the architectural class of Ben van Berkel andJohan Bettum. Since 2002 he has held a tenured professorship at theUniversity of Kaiserslautern, Germany, where he chairs the Department ofStructural Design at the Faculty of Architecture, and where he is building upa research program in the field of “materials and form in architecture.”

http://www.o-s-d.com

Page 263: B. Kolarevic-Performative Architecture

256

DAVID LEATHERBARROWProfessor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

David Leatherbarrow is Professor of Architecture and Chairman of thePhD program at the University of Pennsylvania, where he has taughtarchitectural design and theory since 1984, and where he wasDepartmental Chair between 1992 and 1998. Before going to Penn hetaught in England at the University of Cambridge and at the Polytechnicof Central London. He has also visited and taught at many universities inthe USA and abroad. David Leatherbarrow studied architecture at theUniversity of Kentucky, where he earned his Bachelor of Architecturedegree, and he completed research for his PhD in Art at the University ofEssex, UK.

In his scholarly work he has published a number of books, the mostrecent being Topographical Stories: Studies in Landscape andArchitecture (2004). Surface Architecture (MIT Press, 2002), written incollaboration with Mohsen Mostafavi, won the CICA International BookAward: The Bruno Zevi Prize. Earlier books include Uncommon Ground:Architecture, Technology and Topography (MIT Press, 2000), The Rootsof Architectural Invention: Site, Enclosure and Materials (CambridgeUniversity Press, 1993), and, also with Mostafavi, On Weathering: theLife of Buildings in Time (MIT Press, 1993), which won the 1995International Book Award in architectural theory from the AmericanInstitute of Architects. These books have been frequently reviewed inscholarly journals. Additionally, he has published over fifty scholarlyarticles in architectural journals, including AA Files, ArchitecturalDesign, Center, Daidalos, Journal of Garden History, Journal of theSociety of Architectural Historians, Rassegna, Via, and others.

His awards include a Visiting Scholars Fellowship at the CentreCanadien d’Architecture, the Cass Gilbert Distinguished Professorship atthe University of Minnesota, and a Fulbright Hays Scholarship for studyin Great Britain. In the past, his research has focused on topics in thehistory and theory of architecture, gardens and the city; more recentlyhis work has concentrated on the impact of contemporary technology onarchitecture.

http://www.design.upenn.edu

BRANKO KOLAREVICAssociate Professor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

Branko Kolarevic joined the University of Pennsylvania in1999, where he teaches design and digital media courses,such as “Digital Morphogenesis” and “Digital Fabrication.”Prior to joining Penn, he taught at universities throughoutNorth America (Boston, Los Angeles, Miami) and Asia (HongKong). He has lectured worldwide on digital media in design,most recently on the “virtual design studio,” “relations-baseddesign” and “digital architectures.” In 2000, he founded theDigital Design Research Lab (DDRL) at Penn.

He has published extensively in the proceedings ofACADIA, CAADRIA and SIGRADI, and has edited andauthored several books, including Architecture in the DigitalAge: Design and Manufacturing published by Spon Press,London. He is the Review Editor in Architecture for theAutomation in Construction journal, published by ElsevierScience Publishers, Amsterdam. He is the past President ofthe Association for Computer Aided Design in Architecture(ACADIA). In 1998 he chaired ACADIA’s organizingcommittee of the first Internet-based design competition forthe “Library for the Information Age.”

He received Doctor of Design (1993) and Master inDesign Studies (1989) degrees at Harvard University,Graduate School of Design. He also holds the DiplomaEngineer of Architecture degree (1986) from the Universityof Belgrade, Faculty of Architecture.

http://www.gsfa.upenn.edu/ddrl/

Page 264: B. Kolarevic-Performative Architecture

257

PETER McCLEARYProfessor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

Professor McCleary studied and graduated in applied mathematicsand natural philosophy at Glasgow University, civil engineering atthe University of Strathclyde, structural engineering at ImperialCollege, London, and architecture at the University of Pennsylvania.

He has practiced engineering and architecture in London withOve Arup, Arup Associates, Felix Samuely (with Frank Newby) andFreeman-Fox. He has had an architectural and engineeringconsulting practice since 1965. Among those architects andengineers that he has consulted with are: Lewis Davis, BernardHuet, Louis Kahn, Jean-Marc Lamuniere, Ian McHarg, and RobertLe Ricolais.

Professor McCleary was invited to the University ofPennsylvania in 1965, where he served as Chairman of the Mastersand PhD programs in Architecture, and he was the founder and firstChairman of the Historic Preservation program. He teachesstructures, aesthetics of bridges, philosophy of technology, therelationship between structure and space, and design studio.

He is an invited Professor of Technology and Design at manyuniversities in the US, Canada, Europe, Australia and the MiddleEast. He has received research grants from NEA, the NationalTrust for Historic Preservation, the French Ministry of Culture, andFellowships from NEA and the Fulbright and Graham Foundations.He was the founder of the ACSA Annual Technology Conference,and was awarded their Distinguished Professor Medal.

Professor McCleary continues to write many research andinterpretive articles on the relationship between the technologies ofarchitecture and engineering, which are published in severallanguages in numerous journals and books.

http://www.design.upenn.edu

ALI MALKAWIAssociate Professor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

Ali Malkawi joined the University of Pennsylvania in 2001 asan Associate Professor. He teaches architectural technology andcomputation in the Master of Architecture program. He alsoconducts research in the areas of computational simulation andbuilding performance evaluation. In addition, he supervisesresearch and teaches in the PhD program. He was the founderand former coordinator of the Augmented Reality Group at theUniversity of Michigan, and he currently heads the BuildingSimulation Group at Penn. He taught and conducted researchat the Georgia Institute of Technology, the University ofMichigan and Harvard. He has lectured at numerousuniversities and conferences, and has been published in manyproceedings and journals.

Dr. Malkawi received his PhD in Architecture andArtificial Intelligence in 1994 from the Georgia Institute ofTechnology. He also holds a MArch degree from the Universityof Colorado and a BS in Architecture Engineering from theJordan University of Science and Technology.

http://pobox.upenn.edu/~malkawi

Page 265: B. Kolarevic-Performative Architecture

258

ALI RAHIMAssistant Professor of ArchitectureUniversity of PennsylvaniaPhiladelphia, USA

Ali Rahim is an Assistant Professor of Architecture in Mediaand Design at the University of Pennsylvania and he isDirector of the Contemporary Architecture Practice in NewYork City. He has established an award-winning profile infuturistic work using digital design and production techniques.

He was the recipient of the Honor Award for Excellencein Design from Columbia University where he received hisMaster of Architecture. Current design research includes arange of residential, commercial and product design projects.His architectural designs have received awards and have beenexhibited widely. They include a winning competition entry forEphemeral Structures for the 2004 Olympic Games, ashopping mall, steel museum and a one-acre naval memorial.Past exhibition venues include the Artists Space Gallery inNew York City, and the Royal Institute of British Architects(RIBA) in London. His most recent exhibition was in Delft, theNetherlands. He has also lectured extensively about his workin Asia, Europe and North America.

His guest edited books and journals includeContemporary Techniques in Architecture (Academy Editions/Wiley and Sons, February 2002) and Contemporary Processesin Architecture (Academy Editions/Wiley and Sons, August2000). His projects and articles have been published by ActarPress (Barcelona), Columbia University Press (New York City)and MIT Press (Cambridge, Mass.), among others. Currentlyhe is working on a monograph of his work, Manual for DigitalMedia and Architecture Design (Academy Editions, London,forthcoming 2004), and a special issue for ArchitecturalDesign entitled “Contemporary Cultural Production inArchitecture” (Academy Editions/Wiley and Sons,forthcoming 2005).

http://www.c-a-p.net

MAHADEV RAMANPrincipalArupNew York, USA

Mahadev Raman leads Arup’s Building Engineering Group inNew York and is a member of the firm’s Americas Board. Hehas been with Arup since 1978, providing engineering designleadership for multidisciplinary teams on a wide variety ofprojects worldwide. Prior to joining the New York office, heworked in Arup’s London and Cambridge offices.

He brings particular expertise in the design of sustainable,high performance and energy-efficient buildings. He haspioneered the use of sophisticated analytical techniques toimprove the performance of low-energy designs.

Mahadev Raman lectures regularly at Princeton andColumbia Universities on sustainable design and environmentalcontrol. He has a Bachelor of Engineering Science from theUniversity of Durham, UK (1978) and a Master of Science inApplied Energy from the Cranfield Institute of Technology, UK.

http://www.arup.com

Page 266: B. Kolarevic-Performative Architecture

259

CRAIG SCHWITTERPartnerBuro HappoldNew York, USA

Craig Schwitter is one of the youngest partners in BuroHappold. He established the New York office in 1998 aftergaining overseas experience working on the MillenniumDome in London and various other projects in the UK. Hehas led both the structures and MEP groups on key projects,such as Santiago City of Culture, Rensselaer PolytechnicInstitute and the new green Genzyme Headquarters.

His experience in the Buro Happold UK officesmotivated him to seek to bring the best aspects of theEuropean collaborative team design approach to the UnitedStates. In setting up the New York office, his vision is toutilize the fullest depth of Buro Happold’s internationalpractice knowledge while recognizing the importance ofmobilizing locally-based experience to ensure a project’ssuccess. Under Craig Schwitter’s guidance, and along withhis belief that excellent architecture and excellentengineering are inseparable, the New York office quicklyearned a reputation for outstanding, creative engineering.As a result, the office expanded rapidly to the currentstaffing level of around thirty people.

In 2002, he was appointed as a Trustee for the VanAlen Institute, which is committed to improving the designof the public realm.

http://www.burohappold.com

LARS SPUYBROEKPrincipalNOX ArchitectsRotterdam, Netherlands

Since the early 1990s Lars Spuybroek has been involved inresearching the relationship between architecture and media, oftenmore specifically between architecture and computing. He was theeditor and publisher of one of the first magazines in a book format(NOX, and later Forum), he made video (Soft City) and interactiveelectronic artworks (Soft Site, edit Spline, deep Surface), with thelast five years spent focusing more on architecture (HtwoOexpo,V2_lab, wetGRID, D-Tower, Son-O-house, Maison Folie).

His work has won several prizes and has been exhibited allover the world, including presentations at the Venice Biennale in2000 and 2002. He is currently working on an interactive tower forthe Dutch city of Doetinchem (D-Tower), “a house where soundslive” (Son-O-house), an interactive office building in Stratford-upon-Avon, England (SoftOffice), a multiple complex of culturalbuildings in Lille, France, and a Home for Alice, in Italy.

Lars Spuybroek has lectured all over the world and has taughtat several universities in the Netherlands. He is a regular VisitingProfessor at Columbia University. Since 2002, he has held atenured professorship at the University of Kassel in Germany wherehe chairs the CAD/digital design techniques department. His bookMachining Architecture will be published by Thames & Hudson in2004.

http://www.noxarch.com

Page 267: B. Kolarevic-Performative Architecture

260

ANDREW WHALLEYDirectorGrimshawLondon and New York

Andrew Whalley has worked at Grimshaw since 1986. Heworks across a broad range of sectors and has undertakenprojects in a number of countries which include: DonaldDanforth Plant Science Center in St. Louis, USA; SouthernCross Station, Melbourne, Australia; Fundacion CaixaGalicia, Coruna, Spain; Paddington Station, London; theEden Project, Cornwall, UK; the Royal College of Art inLondon; and the new spa in the city of Bath, UK.

In addition to working on projects, he holds theresponsibility in the office for developing new business andfor overseeing publicity and publications. He also acts as avisiting lecturer at the Bartlett, University College London,and as a visiting professor at Washington University in St.Louis. He has recently spoken at events and symposia inItaly, the USA, Norway, Germany, Holland, and at variouslocations in the UK.

Andrew Whalley studied at the Mackintosh School ofArchitecture in Glasgow and at the Architectural Associationin London. He is registered as an architect in the UnitedKingdom and in the United States.

Andrew also works in partnership with his wife, FionaGalbraith, and together they have designed a house in Dollar,which won an RIBA Award for Scotland, and more recentlythey have completed a house and studio for themselves inLondon. He is based principally in Grimshaw's New Yorkoffice and is overseeing the recent competition-winingproject, the Fulton Street transit center. Andrew Whalley haswritten a book with Hugh Pearman entitled The Architectureof Eden, which was published by Transworld in October2003.

http://www.grimshaw-architects.com

Page 268: B. Kolarevic-Performative Architecture

261

PHOTO CREDITS

COVER Peter Macapia

11.1–14 Anna Catherine Vortmann

Hakes, except:1.6 David Leatherbarrow1.11 David Leatherbarrow

22.1 Michael Dyer Assoc.2.2 Michael Dyer Assoc.2.3 Grimshaw2.4 Jo Reid and John Peck2.5a Andrew Whalley/Grimshaw2.5b Peter Stroebel2.6 Peter Cook/VIEW2.7 Peter Stroebel2.8 Jo Reid and John Peck2.9 Werner Huthmacher2.10 Jens Willebrand2.11a Andrew Southall2.11b Edmund Sumner2.12 Peter Cook/VIEW2.13 Grimshaw2.14 Grimshaw2.15 Michael Dyer Assoc.2.16 Timothy Hursley2.17 Grimshaw2.18 Grimshaw2.19 Jens Willebrand2.20 Nathan Willock2.21 Grimshaw2.22 Edmund Sumner2.23 Grimshaw2.24 Grimshaw2.25 Edmund Sumner2.26 Grimshaw2.27 Grimshaw2.28a Grimshaw2.28b Grimshaw2.29a Grimshaw2.29b Grimshaw/Univ. of Bath/

UCL/Bentley2.29c Grimshaw/Univ. of Bath/

UCL/Bentley2.30 Grimshaw2.31a Grimshaw2.31b Grimshaw

33.1–17 Arup, except:3.2 BNIM Architects3.5 Andy Ryan3.11 Frederick Charles3.12 Frederick Charles3.13 Peter Mauss/ESTO3.14 Peter Mauss/ESTO3.17 Yutaka Nagata © United

Nations UN/DPI

44.1 Architectural Record/McGraw-Hill4.2 Productivity Press4.3 Productivity Press4.4 Greg Lynn/Princeton Architectural

Press4.5 Regents of the Univ. of California4.6 Reinhold Publishing Corp.4.7 Sweets Construction/McGraw-Hill4.8 Sweets Construction/McGraw-Hill4.9 Ottomar Gottschalk/Verlag Schnelle4.10 Ottomar Gottschalk/Verlag Schnelle4.11 Ottomar Gottschalk/Verlag Schnelle4.12 Architectural Review4.13 Van Nostrand Reinhold Company4.14 Architectural Record/McGraw-Hill,

Progressive Architecture4.15 Ante Prima, Paris/Birkhauser, Basel

55.1–34 Herzog + Partner, except:5.3 Peter Bonfig5.4 Klaus Kinold5.7 Dieter Leistner5.8 Dieter Leistner5.9 Verena Herzog-Loibl5.10 Dieter Leistner5.11 Dieter Leistner5.12 Peter Bartenbach5.13 Verena Herzog-Loibl5.15 Dieter Leistner5.16 Verena Herzog-Loibl5.18 Verena Herzog-Loibl5.19 Verena Herzog-Loibl5.20 Reinhard Demuss5.22 Dieter Leistner5.23 Dieter Leistner

5.25 Dieter Leistner5.27 Dieter Leistner5.28 Dieter Leistner5.31 Dieter Leistner5.34 Moritz Korn

66.1–7 Ali Malkawi, except:6.1 Glenn Forney/NIST6.5 IAI

77.1–7 Godfried Augenbroe

88.1–25 Buro Happold, except:8.10 Buro Happold/Adam Wilson8.17 Foster and Partners/Nigel Forster8.18 Foster and Partners/Nigel Forster8.20 Mandy Reynolds/Fotoforum8.21 Mandy Reynolds/Fotoforum8.22 Mandy Reynolds/Fotoforum

99.1–12 RFR

1010.1 Harald Kloft10.2 Harald Kloft10.3 Irving Penn © Austrian Frederick

and Lillian Kiesler PrivateFoundation

10.4 Fritz Busam10.5 ABB/Bernhard Franken10.6 ABB/Bernhard Franken10.7 ABB/Bernhard Franken10.8 Bollinger + Grohmann10.9 Bollinger + Grohmann10.10 Harry Schiffer10.11 spacelab.uk10.12 Bollinger + Grohmann10.13 Bollinger + Grohmann10.14 Bollinger + Grohmann10.15 Bollinger + Grohmann10.16 Bollinger + Grohmann10.17 Bollinger + Grohmann10.18 Bollinger + Grohmann10.19 Franken Architekten

Page 269: B. Kolarevic-Performative Architecture

262

10.20a osd – office for structural design10.20b osd – office for structural design10.21 ABB/Bernhard Franken10.22 Harald Kloft10.23 Bollinger + Grohmann10.24 Harald Kloft10.25a Harald Kloft10.25b Harald Kloft10.25c Harald Kloft10.26 Harald Kloft10.27 Fritz Busam10.28 Tilman Elsner

1111.1–16 realities:united, except:11.1 Harry Schiffer11.2 Colin Fournier11.3 Colin Fournier11.6 ArGe Kunsthaus, Graz/

Pichlerwerke GmbH, Graz11.7 Harry Schiffer11.11 Landesmuseum Joanneum, Graz

11.12 Harry Schiffer11.16 Landesmuseum Joanneum, Graz

1212.1–9 NOX/Lars Spuybroek

1313.1–9 Ali Rahim/Contemporary

Architecture Practice

1414.1 Bernhard Franken/ABB Architekten14.2 Bernhard Franken/ABB Architekten14.3 Arup14.4 Future Systems14.5 Bollinger + Grohmann14.6 Arup14.7 Gehry Partners14.8 Kristina Shea14.9 Kristina Shea14.10 Matthew Herman/Univ. of

Pennsylvania

1515.1 NOX/Lars Spuybroek15.2 NOX/Lars Spuybroek15.3 NOX/Lars Spuybroek15.4 NOX/Lars Spuybroek15.5 Colin Fournier15.6 Landesmuseum Joanneum, Graz15.7 Graeme Peacock15.8 Gifford and Partners15.9 Colleen and Ted Garringer15.10 Alan Karchmer/ESTO15.11 Mark Burry and

Mark Goulthorpe/dECOi15.12 Foster and Partners15.13 Renzo Piano Building Workshop15.14 Foster and Partners15.15 Arup

1616.1–3 Arup

Page 270: B. Kolarevic-Performative Architecture

263

INDEX

AABACUS, 196ABB Architekten, 138–147, 195, 196

Dynaform, 138–147, 195, 196acoustic

analysis, 197performance, 38simulation, 49, 130

acrylic glass, 143, 153, 156, 207, 212active systems, 77adaptability, 23, 232air conditioning, 46, 52, 53, 57, 64–66, 222Air Diffusion Performance Index (ADPI), 106airflow simulation, 80Aish, Robert, 233, 234Allies and Morrison, 119

BBC White City, London, 119analog computing, 167, 168, 170, 171, 210animation, 44, 50, 59, 139, 179, 182, 198Archigram, 138, 153, 205architectural performance (see also

performance), 7, 8, 99, 205Aristotle, 8, 17, 18, 241artificial intelligence, 90Arup, 43–54, 196, 197, 211, 212, 217–223

Loughborough Univ. of Tech. Lab, 223Mining and Metallurgy Labs, 217Sydney Opera House, 217, 220

Arup, Ove (see Arup)ASHRAE, 106ASTM, 99, 100Ateliers Jean Nouvel (see Nouvel, Jean)Augenbroe, Godfried, 97, 237–245, 253augmented environments, 92

BBan, Shigeru, 117

Japanese Pavilion, Hanover, 117Baudrillard, Jean, 7Behnisch, Günter, 137

Olympic Stadium, Munich, 137Bentley Systems, 38biomimicry, 57biomorphic structure, 137, 153, 205, 212biotechniques, 57–70Bilbao effect, 209BIX communicative display, 151–160, 207Blassel, Jean-François, 123, 234–240, 253BLAST, 87blob modeling, 59BNIM Architects, 45

University of Texas Nursing School, 45Bollinger + Grohman, 196, 197Braham, William, 55, 225–233, 254Breuer, Marcel, 9Bruder, Will, 9

Phoenix Central Library, 9

buildingcodes, 43, 99engineering, 119, 122, 234maintenance, 103–105performance (see also performance),

3, 43, 87, 90, 100, 104–108,127, 210, 212, 249

regulations, 43, 99, 144serviceability, 99, 100simulation (see also simulation), 88,

91, 115, 239, 241, 242, 249standards, 144systems, 101, 102, 119

Building Design Advisor (BDA), 89Buro Happold, 115–121, 239Bürolandschaft, 63, 64, 66, 165

CCAD, 36, 38, 115, 119, 140, 196CAM, 36, 38, 113, 115, 119, 120Calatrava, Santiago, 208, 210

Milwaukee Art Museum, 208Candela, Felix, 137CAP (Contemporary Architecture Practice,

see Rahim, Ali)CATIA, 140Centre Pompidou, 68, 137CFD, 47, 52, 87–89, 91–93, 113, 118,

119, 196, 197, 239, 240, 242chaos theory, 222Chareau, Pierre, 12

Maison d’Alsace, 12Chaszar, Andre, 236Christoph Ingenhoven Overdiek Arch., 116

Stuttgart 21 Train Station, 116Chrysler Building, 53cladding, 23, 27, 32, 117, 126classicism, 220climate modeling, 58CNC (Computer Numeric Control), 206code validation, 87commissioning, 50composite materials, 210computational

fluid dynamics (see CFD)instruments, 249modeling, 120simulation (see simulation)tools, 113–115, 122, 139, 239

computer-aided design (see CAD)computer-aided manufacturing (see CAM)collaborative environments, 88constructivism, 168, 174, 175Cook, Peter,

Kunsthaus Graz, 140, 141, 143, 147,151–160, 197, 207, 209, 212

cooling, 47, 52, 73, 82, 102, 132, 222, 236

Corbusier (see Le Corbusier)Croxton, Randolph, 53Crystal Palace, 77cybernetics, 58, 217, 221, 223

Ddata

exchange standards, 89integration, 88model, 89, 90representation, 92sharing, 89, 90visualization, 92

decision-support environments, 90, 91deconstructivism, 174, 220Deleuze, Gilles, 210Descartes, 217, 219design

analysis, 88, 90–92evaluation, 91, 92evolution, 92, 114methods, 217optimization, 91representation, 91space, 92synthesis, 91

determinism, 57digital simulation (see simulation)display technology, 152, 155DOE (United States Dept. of Energy), 87double-glass façade, 68, 78, 84double-skin façade, 30, 82, 89, 118, 132Drew, Jane, 9Duffy, Francis, 64Duthilleul, J. M., 126

TGV train station, Avignon, 126–128dynamic architecture, 24

EEco-Tech (architecture), 211Edler, Jan, 149, 237, 245, 254eifForm, 199electrochromic glass, 74emergency evacuation simulation, 50Empire State Building, 53energy

balance, 73, 77codes, 127conservation, 65, 66, 81consumption, 43, 73, 74, 78, 82, 107modeling, 119preservation, 44

Energy-Star labeling, 108EnergyPlus, 87, 88environmental

comfort, 67conditioning, 67, 68

Page 271: B. Kolarevic-Performative Architecture

264

design, 84effects, 58, 65, 119energy, 81illness, 67performance, 57, 210–212pollution, 74problems, 59simulation (see also simulation), 87systems, 113

Environmentally Viable Architecture(EVA), 38

ergonomics, 219ETFE foil, 33–35evacuation simulation (see also people flow

modeling), 50events (in architecture), 10–14evolutionary techniques, 91EXPRESS, 104

Ffabrication, 116, 119, 120feedback, 58, 65, 114, 186, 234Feichtinger Architects, 129

Bercy footbridge, 129fiberglass paneling, 23finite elements, 115, 129, 142, 143, 196,

197, 235fire

engineering, 51simulation, 88

Fire Dynamics Simulator (FDS), 88Flemming, Ulrich, 198flexibility (of use), 23, 25, 81, 163fluid mechanics, 88form

finding, 138, 140, 144, 174, 195, 199generation, 92, 137–143, 147, 195–199

Forrester, Jay, 58–60Foster and Partners (see Foster, Norman)Foster, Norman, 49, 68, 120, 196, 210–212

Commerzbank, 68Gateshead Music Centre, 120Great Court, British Museum, 120, 121Greater London Assembly (City Hall),

49, 197, 211, 212Swiss Re, 196, 211

Fournier, ColinKunsthaus Graz, 140, 141, 143, 147,

151–160, 197, 207, 209, 212Frank O. Gehry & Assoc. (see Gehry, Frank)Franken Architekten (see Franken, Bernhard)Franken, Bernhard, 138–147, 195, 210, 231

Dynaform, 138–147, 195, 196“Klöpp” complex, Reykjavik, 143, 147

freeform architecture, 137, 138, 142, 147,174, 229

Friedmann, Yona, 74

Fry, Max, 9FTL Design Engineering Studio, 115

Darien Lake Pavilion, 115Fuller, Buckminster, 57functional

performance, 28requirements, 101, 102

functionalism, 8, 9, 57, 163Future Systems, 196, 197

Project ZED, 196, 197

Gg-value, 73GA (see genetic algorithms)Gallacio, Anya, 30Gaudí, Antoni, 167, 210

Sagrada Familia Church, 167Gaussian

analysis, 197curvature, 222

Geddes, Patrick, 57Gehry, Frank, 9, 140–143, 197, 209, 241

Guggenheim Museum, 209Experience Music Project, 9, 197M.art.A. Museum, Herford, 140, 142

Gehry Partners (see Gehry, Frank)General AEC Reference Model, 100General Systems Theory, 58, 100, 217generative design, 91, 195, 199, 200genetic algorithms, 91, 92Gensler, 47

GAP Headquarters, 47Gielingh, Wim, 100Gifford and Partners, 207Glaspalast (Munich), 77global warming, 43Gothic architecture, 210Gould, Stephen J., 179Goulthorpe, Mark, 207, 209, 210

Aegis Hyposurface, 207, 209gradient-based methods, 91grass roof, 46green design, 65, 66Grimshaw (see Grimshaw Nicholas)Grimshaw, Nicholas, 23–40, 118, 119, 139,

210, 211, 227, 232Battersea Power Station, 30British Pavilion, Expo ’92, 27, 227Concert Hall, RPI, 38, 39, 118, 119Danforth Plant Science Center, 31Eden Project, 34–36Exhibition Hall (Frankfurt), 33, 139Financial Times Building, 24, 25Fundacion Caixa Galicia, 30, 31HAL9000, 32, 33Herman Miller Factory (Bath), 23, 24Herman Miller Warehouse, 23, 24

IGUS Factory, 24Ludwig Erhard Haus, 28National Space Centre, 33Paddington Station, 29Royal College of Art, 31, 32Southern Cross Railway Station, 36V&A Museum Boilerhouse Extension, 30Waterloo Station Terminal, 25–27

Gruen, Victor, 60Guattari, Felix, 210

Hhabitability, 101Hardy, G. H., 220heat

energy flows, 88exchange unit, 82loss, 80, 212recovery, 77, 80transfer, 220

heating, 74, 76, 77, 82, 102, 103, 222Heidegger, Martin, 218, 219, 223Herzog, Thomas, 71–84, 210, 255

Design Center, Linz, 77–80DMAG Administration Building, 81–84Guest Building, Windberg, 75–77

Herzog + Partner (see Herzog, Thomas)high-performance building, 249High-Tech (architecture), 211HOK Sport, 120, 122

Arsenal Stadium, London, 120, 122Holl, Steven, 47, 48

Residence Hall, MIT, 47, 48holographic glass, 31Honzik, Karel, 57human health, 57, 68Husserl, Edmund, 218, 219HVAC, 106–108, 113–115, 119

Iimmersive building simulation, 92, 93indoor air-quality assessment, 88Industry Foundation Classes (IFC), 89–90information flow, 60–62instrumentality, 18integrated building

components, 68design, 44, 115models, 89

intelligent building systems, 74interactive

environment, 164façade, 152system, 205

International Alliance of Interoperability, 89International Organization for Standardization

(ISO), 101

Page 272: B. Kolarevic-Performative Architecture

265

interoperability, 88–90, 93, 241Isler, Heinz, 138

JJourda and Perraudin, 210

KKahn, Louis, 14, 68, 172, 217, 219

Margaret Esherick House, 14Richards Medical Labs, 68

Kiesler, Frederick, 57, 68, 137Endless House, 137

kinetic architecture, 24, 208Kloft, Harald, 135, 225, 229–232, 236, 255knowledge-based systems, 90Kohn Pederson Fox Associates, 107

Sam Nunn Atlanta Federal Center, 107Kolarevic, Branko, 1, 193, 203, 225, 227–

233, 237, 240–246, 256Kroeber, A.L., 60Kurokawa, Kisho, 74

LLarge Eddy Simulation, 88Le Chatelier (principle of Le Chatelier), 220Le Corbusier, 12, 18, 57, 68, 218, 233

Carpenter Center for Visual Arts, 18Palace of the Soviets, 12

Le Ricolais, Robert, 217Leadership in Energy and Environmental

Design (see LEED rating)Leatherbarrow, David, 5, 225, 227–229, 231–

233, 235, 239, 246, 256LEED rating, 44, 46, 53, 103Legorreta + Legorreta, 17

Santa Fe Art Institute, 17light pollution model, 119lighting

performance, 91simulation, 49, 119

lightweight structure, 35Lloyd Wright, Frank, 68, 218Lotka, Alfred J., 220low-embodied energy construction, 36low-energy

building systems design, 119devices, 132

Lönberg-Holm, Knud, 61, 62, 68Lynn, Greg, 59, 60, 195, 210

MMach, Ernst, 219Macintosh, Charles, 31

Glasgow Art School, 31Mahdavi, Ardeshir, 198Malkawi, Ali, 85, 225, 231–235, 237, 239–

243, 246, 247, 257

manufacturing (see also fabrication and CAM),26, 138, 141, 144, 147, 174, 186, 190

Maver, Thomas, 196, 198Maya animation software, 139, 143McCleary, Peter, 215, 237, 239, 241, 244–

246, 257McLuhan, Marshall, 207Mean Radiant Temperature (MRT), 106mechanical systems (see also HVAC), 114, 142media

façade, 152, 153, 159, 207, 212installation, 153, 156, 207

Meier, Richard, 15, 51The Getty Center, 15Federal Courthouse, Phoenix, 51

Merleau-Ponty, Maurice, 7mixed-mode (ventilation) system, 48modernism, 9, 163, 220modular elements, 24Moeding façade system, 82monocoque structure, 210morph visualization, 92morphing, 92morphogenetic (design), 58, 59, 195, 200morphology, 195, 210, 220movement (in architecture), 12–15Mumford, Louis, 57Murcutt, Glenn, 210

Nnatural ventilation (see also ventilation), 46,

47, 81, 82, 88, 211Navier-Stokes equations, 88Nervi, Pier Luigi, 137

Palazetto dello Sport, 137noise prediction, 88Nolan, Thomas, 60non-linear

analytical software, 186design (process), 47, 114finite elements programs, 115

non-standard architecture, 137, 147, 174Nouvel, Jean, 12, 16

L’Institut du Monde Arabe, 12NOX (see Spuybroek, Lars)

Oobject-oriented environment, 88, 198Olgay, Aldar and Victor, 9operations research, 221optimization, 84, 90–93, 113, 114, 138, 142–

147, 186, 190, 196, 199, 212, 229, 236organic architecture, 27, 59organicism, 63Ortega y Gasset, Jose, 218, 219, 229Otto, Frei, 137, 167, 170–174, 199, 210Otto, Greg, 237, 242, 246

PPakesch, Peter, 159parametric

design, 126, 139, 147modeling, 120, 147

passive systems, 77people flow modeling (see also evacuation

simulation), 120, 122performance

analysis, 90, 108, 198appraisal aids, 196assessment, 105criteria, 23, 36, 38, 210, 222evaluation, 92, 102, 198, 199indicator (PI), 100–109measures, 108methods, 90metrics, 100, 103, 104modeling, 198quantification, 100, 103requirements, 101–102scoring, 100simulation (see also performance-based

simulation), 87, 90–92techniques, 90tracking, 104variation, 187

performance-basedbuilding method, 99decision-making, 90design, 3, 30, 87, 92, 113–122, 137,195, 196, 199, 200engineering design, 113requirements, 99simulation, 3, 90–93, 195, 197, 227structural design, 115

performance-driven design (see performance-based design)

performativity, 153, 179, 182, 186, 190, 225–230, 232, 233, 236, 245, 246

photovoltaiccells, 28, 197systems, 74

Piano, Renzo, 12, 210, 211Aurora Place, 12Tjibaou Cultural Center, 211

Polshek Partnership Architects, 118, 119Biomedical Research Building, 118, 119

postmodernism, 246post-occupancy evaluation (POE), 105pragmatism, 23Predicted Mean Vote (PMV) distrib., 103, 106prediction accuracy, 87prescriptive design, 113–115probabilistic inversion, 243problem-solving, 90, 91, 125process-driven integration, 88, 90

Page 273: B. Kolarevic-Performative Architecture

266

Prouve, Jean, 24Pye, William, 28

RRahim, Ali, 177, 180–191, 225–229, 235, 258

Catalytic Furnishings, 186–191Leisure Generator, 180, 181Variations, 181–185

Rajchman, John, 205Raman, Mahadev, 41, 225–236, 258rapid prototyping, 36, 38realities:united, 151, 153, 155, 159, 207RFR Consulting Engineers, 125–133

Bandshell, 130Bercy Footbridge, 129Strasbourg Train Station, 131, 132TGV train station, Avignon, 126–128

Rhinoceros, 141Roche & Levaux, 68

[Un]plug Building, 68Rogers, Richard, 236Romanesque architecture, 210Rossi, Aldo, 8

SSaarinen, Eero, 137

TWA Terminal, New York, 137Sasaki Associates, 45

American University, Cairo, 45Sauerbruch and Hutton, 210Schuller, Matthias, 131Schulze-Fielitz, Eckhard, 74Schwitter, Craig, 111, 234, 235, 259Seagram Building, 65semi-monocoque structure, 210SEMPER, 89, 198Semper, Gottfried, 174Sert, José Luis, 9Serviceability Tools and Methods, 99, 100, 105Shea, Kristina, 199sick building syndrome (SBS), 66–68Simon, Herbert, 217, 223simulated annealing, 91, 199simulation, 3, 44, 49, 50, 59, 73, 74, 80, 87–

93, 115, 119, 120, 125, 129–131, 195–198, 212, 227, 234–236, 239–242, 249algorithms, 87engine, 87, 88environment, 87, 88, 198tools, 87, 90, 91

Simulex, 50Skidmore Owings & Merrill, 48–50, 116–117

Penn Station, New York, 48, 49Roppongi Canopy, Tokyo, 116–117Terminal 4, JFK Airport, 50

Slessor, Catherine, 211smoke visualization, 88

SmokeView, 88solar

collectors, 77cooling system, 73design, 73diagram, 210energy, 73, 76, 77, 131harvesting, 46heat gain, 46, 47, 73, 78, 212heating, 73, 76radiation, 76, 220storage units, 77

Solomon Cordwell Buenz & Associates, 45space planning, 219spacelab.uk (see also Cook, Peter), 153, 207Speaks, Michael, 59Spuybroek, Lars, 164–169, 171–173, 205–

207, 210, 237, 239, 241–246, 259D-Tower, 205, 206Maison Folie, 205, 207obliqueWTC, 171–173SoftOffice, 164–168

structuralanalysis, 116behavior, 7, 129, 130, 141, 220engineering, 115, 137, 142, 143, 199optimization, 138, 143, 147performance, 142, 197, 210shape grammar, 199system, 113, 114, 117, 141–143, 197

Sullivan, Louis, 65Wainwright Building, 65

Sutnar, Ladislav, 61sustainability, 27, 43, 101, 195, 205sustainable

architecture, 38building, 81, 125, 212design, 43–45

Suuronen, Mati, 137Sweets Catalog, 60–62

TTange, Kenzo, 74technological sustainability, 158tensile fabric structure, 115TENSYL, 115, 116Terragni, Giuseppe, 12

Casa del Fascio, 12theory of complexity, 220thermal

comfort, 102–106, 230flows, 87radiation, 106performance, 91, 92, 127simulation, 52, 119, 131storage mass, 82transmission, 84

thermo-active floor slabs, 82thermochromic glass, 74Thom, Rene, 222threshold effects, 59, 60, 62, 64–66time-based modeling, 139, 195Tschumi, Bernard, 51

Lerner Student Center, 51Tsien, Billie, 7, 13

Neurosciences Institute, 7, 13

UU-value, 73, 74, 106United Nations Headquarters, 53

Vvacuum thermal insulation, 74vacuum-tube collectors, 77van der Rohe, Mies, 14

Tugendhat House, 14Wolf House, 14

vector analysis, 220ventilation (see also natural ventilation), 24,

58, 66, 67, 73, 74, 80efficiency, 92performance criteria, 91solution, 118, 119

Venturi effect, 80Venturi, Robert, 174verification methods, 101Vierendeel system, 144Virilio, Paul, 172virtual environments, 92, 139, 140visualization, 44, 87, 90, 91, 93, 197

Wwater recycling, 46Whalley, Andrew (see also Grimshaw), 21,

225, 227, 228, 230–234, 260Wilkinson Eyre Architects, 208

Millenium Bridge, Gateshead, 208William McDonough + Partners, 47

GAP Headquarters, 47Williams, Tod (and Billie Tsien), 7, 13

Neurosciences Institute, 7, 13wind energy, 82wind-tunnel testing, 80, 212

ZZumthor, Peter, 14

0–93D

modeling, 141, 147printing, 119scanning, 140simulation, 158

4D modeling, 139