Generated Lamella - CumInCAD

340ACADIA2010life in:formationauthor:organization:country:Martin Tamke, Jacob RiiberHauke JungjohannCITA, Centre for Information Technology and Architecture, Royal Danish Academy of Fine Arts, School of ArchitectureKnippers Helbig Advanced EngineeringDenmarkGenerated Lamellaabstract:The hierarchical organization of information is dominant in the setup of tectonic structures. In order to overcome theinherent limitations of these systems, self-organization is proposed as a means for future design. The paper exemplifiesthis within the research project “Lamella Flock”.The research takes its point of departure in the structural abilities of the wooden Zollinger system: a traditional structurallamella system distributed as a woven pattern of interconnected beams. Where the original system has a very limited set ofachievable geometries our research introduces an understanding of beam elements as autonomous entities with sensorymotorbehaviour. By this means freeform structures can be achievedThrough computation and methods of self-organization, the project investigates how to design and build with a systembased on multiple and circular dependencies.Hereby the agent system negotiates between design intent, tectonic needs, and production. The project demonstrates howreal-time interactive modeling can be hybridized with agent–based design strategies and how this environment can belinked to physical production. The use of knowledge embedded into the system as well as the flow of information betweendynamic processes, Finite Element Calculation and machinery was key for linking the speculative with the physical.

ACADIA2010life in:formation341 | 412Figure 1. Detail of the 1:1 Lamella Flock demonstrator.1 Questions of linear processesin architectural designIn a time when relational and generative tools becomecommon in architectural practice and the design ofinformation within projects gains significant attention,the questions that Herbert Simon raised in 1969 in hisbook, “The Sciences of the Artifical,” gain new relevance.Herein he asks how (artificial) systems can be organizedthat can handle complex design tasks and which rolecomputational methods as Artificial Intelligence orGenetic Algorithms have in this process.Since Simons book, a new design practice has emergedin which architects become the developer of bespokedesign environments that allow dynamic interfacingbetween design intention and contextual information(Kolarevic 2005; Schwitter 2005; Burry 2005).Within this practice, the crucial step becomes the designof the system itself, or as Mark Burry calls it “the designof design”. The common approach to dealing with thecomplexity of architectural projects is to separate theminto many levels of subassemblies with hierarchicalinterrelationships. Where in Simons theoretical modelparts in the same hierarchical level can establishinterrelationships, the common practice of parametricdesign leaves them independent. This top-downapproach requires a high amount of specification on alllevels. This causes problems in design projects, whosenature is that huge amounts of required informationare to be developed within the process of the project(Rittel 1973). This insufficiency does not only apply tothe content of information but especially to its relationto other parts. Every designer knows probably exampleswhere information on a specific part clarified in the verylast moment, such as prohibiting fire laws or productionrelated restrictions, caused severe changes and delaysto a project. Where parametric models may representthe geometrical dependency of a part on information ofa higher level, this higher level may yet be affected bythe parts properties. The appearing interdependenciescannot necessarily be defined in linear relationships butrather create networks of relationships with a high levelof complexity. This level increases when parametersof very diverse nature as economical or time basedconstraints are taken into account.Another field of questions concerns the depth ofcomplexity in a project. This can be seen in architectureas the amount of parts in a subassembly. Kieran andTimberlake have shown in 2003 how other industriesgained significantly in productivity by introducingseveral levels of prefabrication. This improved thequality of the final assembly as fewer parts had to becombined and allowed for better quality products asthe parts itself could become more complex. On anarchitectural level, this can be either achieved by morepreassembly or by the use of fewer parts that itselfcan establish more complex relations to their outerenvironment. This would not only reduce complexity onthe parts level but offer chances for more efficiency onall resource levels.As shown in processes of large scale Industries by Kieranand Timberlake, insight and control of all aspects fromdesign to operation raises efficiency immensely. Thereality in architecture appears yet somehow patchy.The downfalls of top-down strategies lead to a wideinterest in alternative design strategies such as bottomupmethods (Kolarevic 2005) or evolutionary design inorder to optimize towards multiple goals of diversenature (Terzidis 2006).paper session | Generated Lamella

342ACADIA2010life in:formationIt becomes obvious that design systems have to bedeveloped that inherit the property to integratecontinuously new knowledge, followed by a renegotiationof a satisfying solution on a certain level, while offeringa flexibility on higher levels to cope with the emergingeffects of this behavior.Within our research we found this complex of questionswhile working on freeform lamella structures. Weintroduce the concept of an aware design model and useself-organization and bottom-up principles as means toallow design in environments that are characterized bymultiple and circular dependencies. We furthermorecombine traditional wood craft with digital informationand fabrication techniques in order to gain efficiencyby higher complexity on part level. The requiredprerequisites are shown in the next two chapters.Figure 2 and 3. Friedrich Zollinger and a originalZollinger Roof system in Merseburg (Germany).2 Zollinger – A Lamella Wood sSystemThe Zollinger construction is a type of Lamella roofconstruction (JS Allen 1999) that was invented in the1920s in order to create wide spanning constructionsout of short pieces of timber (Figure 2 and 3).Figure 4. Principal pattern of the Zollinger lamella structure.The lamellas structural principle consists of acrisscrossing pattern of parallel arches of relativelyshort members. These are hinged together and form aninterlocking network in a diamond pattern. The ingenuityresides within two constituents: the efficient jointsystem that minimizes the amount of shared meetingpoints allowing for simple assembly, and structuralstrength given by the interwoven beams (Figure 4).Where similar systems, as mutually supporting beamsystems (Popovich 2008), usually form barrel or domeshapes work from the AA (Hensel and Menges 2007)and Shigeru Ban (Tristan et al. 2007) demonstratesthe principal ability of the system to form differentshapes using the flex of the material, tolerances in thejoint geometries, and changes in the system’s localorientation. In this bottom-up approach, each elementis threaded individually as it acts autonomously in alarger formation.Our own investigations revealed that freeformstructures can be manually crafted from straightbamboo sticks by exploitation of tolerances in the joint.Yet this method relies purely on skill in crafting andnegotiation with the physical model. The translationFigure 5. Wireframe view of the customized monolithicTenon joints used in the construction.Figure 6. Output of a rule based distribution of lamella elementsGenerated Lamella | paper session

ACADIA2010life in:formation343 | 412of the craft–based approach into an architecturalplanning practice that would allow it to anticipateand fabricate geometry in relevant scale and tolerancebecame a main concern of the investigation.3 Free form wood structuresand previous experiencePrevious research on mass customized parametricwood constructions (Tamke et al. 2008) indicatedthat digital production can provide the sought afterflexible, effective fabrication of easily assembled woodbeams. This approach is based on the conjunctionof computation, digital fabrication, and traditionalcraft techniques. Herein modern CNC wood joinerymachinery allows the cutting of monolithic joints inhigh speed and variable geometry (Figure 5). Thesejoints allow for fast assembly as they incorporate selfregistering geometrical properties such as contemporaryindustrial snap fit joints (Schindler 2009). The improvedunderstanding of forces within massive wood, in itsmonolithic joints as well as in its assembly as structuralsystems through Finite Element systems (Holzner 1999),allows for new applications of traditional wood crafts.The combination of computational capabilities withdigital fabrication therefore allows the introduction ofcraft related knowledge into contemporary practice thatwas previously bound to the skill and knowledge of theexecuting craftsperson.4 Investigating freeform lamella systemsIn the initial stages of the research, the distribution andcomputing of elements were investigated, looking for themost suitable method of controlling the system and thenon-linear relationships within.The lamella structure was at first distributed onpreconceived test surfaces. This presented twoproblems: when following a free-form surface, allbeam endpoints should be on the surface. Since allendpoints also connect to the midpoints of otherbeams, this criterion cannot be met. Secondly, this topdownapproach lacked the possibility of exploring theperformance of the structural principle. How would therigidity of the reciprocal relationship between beamsaffect the scope of shapes possible?The conclusion to use bottom-up approaches instead,gave, at first, problems in controlling the system. Theelements were here structured through rule based lineardistribution where elements were sequentially inserted.Due to the fact that in a networked lamella system,one element is affecting all its neighbors, this resultedin compelling morphologies (Figure 6) but impededdesign control. The linear distribution led to extreme andunpredictable conditions.We could now state the requirements of our system:a bottom-up process with the ability of dynamic nonlinearinteraction where different design possibilitiescould be explored. We introduced an understandingof the structure as a self-organizing system of entitiespossessing a simple set of behavioral properties andrelations to each other.5 An outline of self-organizationTheories of self-organization were originally developedin the context of physics and chemistry. Later it wasfound that these ideas could be extended to thesimulation of social insects. Their colonies solve problemin decentralized systems, comprised of many relativelysimple interacting entities (Bonabeau et al. 1999).This relies on the anti-classical-AI idea that a group ofagents may be able to perform tasks without explicitrepresentations of neither environment nor other agents,and where planning may be replaced by reactivity(Carranza and Coates 2000). By recontextualizing thisinto numerous fields of knowledge, powerful tools fordeveloping dynamic and intelligent systems emerged. Thecontinuous negotiations within such systems are similarto the bespoke conditions needed to process material toachieve material performance within traditional crafts.The advantages reside in flexibility to function, inchanging environments and robustness through theability to function even though some entities may fail toperform. The disadvantages can be located in the bottomupapproach to programming such systems. Here, thepaths to problem solving can never be predefined but arealways emergent and result from interactions amongstentities themselves as well as between entities and theirenvironment. Therefore, using self-organization to solve aproblem requires precise knowledge of both the individualbehaviors and of what interactions are needed to producea desired global effect (Bonabeau et al.1999).paper session | Generated Lamella

344ACADIA2010life in:formation6 The generated lamella system, structure,and behaviourIn order to handle the complexity of the structure, weintroduce a sublevel within the overall structure thatconsists of autonomous elements. These form an innerenvironment, acting on specific inputs from an overallouter environment (Simon 1996). These entities arebased on the interaction of four line segments comingtogether in a spiraling motion. Each entity exhibitswithin itself the non-linear relationship that mostunmistakably defines the global structure it is aimingat. The communication is sematectonic, as relevantinteractions between entities occur only throughmodifications of the environment (Wilson 1975).Initially, the amount of entities, their sizes, and apreliminary distribution of these as a diagonal gridin space are defined. This can either be coherent orfragmented. In both cases, entities are positioned by adistribution of point coordinates or loading a previouslysaved model into the system. This last feature alsoserved for interfacing with other tools (Figure 7, 8).While running, the system is controlled through fourbehavioral algorithms that accumulate vector information(Figure 9). A method that is inspired by the division intogoal types as found in the simulation of flocks, herds, andschools (Flake, 1998). Each algorithm produces directionsand velocities that interact to produce the overallmovement and transformation of an entity:1. Movement towards neighbors: If not representinga corner or an edge, each entity has four neighbors. Bymeasuring the distance and direction from endpoints ofline segments to a neighbor connection point, vectorsare calculated. These vectors are added and weightedto calculate a mean vector by which all points in anentity are moved.Figure 7,8. Renderings of two generatedmodels with diverse characteristicFigure 9. Initial state and the 4 main behavioralprinciples of Lamella FlockFigure 10. Formation and interaction of a lamellastructure within the processing interface2. Orienting towards neighbors: By altering theconfiguration of angles between segments, eachelement tries to orient its segments towards theirneighbors. A segment is in this way sought to be alignedwith the trajectory towards its destination.3. Stretching towards neighbors: Through the aboveorientation, a segment will, within a certain tolerance,be able to stretch to connect to a neighbor. This isallowed when the orientation is correctly aligned andif it is happening within a predefined size limitationof a segment.Figure 11. Processing interface with model of the ROM exhibition siteGenerated Lamella | paper session

ACADIA2010life in:formation345 | 4124. Scale entity: Each entity has the ability to scale upand down while keeping its proportions. This allows fora global push/pull effect within the lamella network.Additionally, production related constraints, such asthe limitation to producible wood joints and the abilityto offset shared beam meeting points, were introducedinto the program. The generative design process wasinformed by its implementation and realization in 1:1.Figure 12. Labeled Non-standard wood beams ready for assemblyThe global behavior occurring from these functionsproduces a network of entities that attempts to obtainthe shape of a surface. The global configuration iscontinuously and non-lineally renegotiated until astable result is achieved.7 A hybrid systemIt seems that already the early work within the use ofAgent based systems in architectural design (Carranzaand Coates 2000) favored an approach that analyzes agiven environment with a set of given parameters in atime-based way and presents the designers solutions.The designer is hereby excluded from the actual designprocess in any other way than setting the initialparameters. The selforganization apparatus becomesa blackbox that offers a design approach alien tosuccessful intuition based iterations of design praxis(Simon 1996) Our experience (Tamke et al 2009) hasshown that in the context of architectural design, acombination of generative and interactive modeling isuseful. We introduced manual manipulation of entitieswhile the system is running. Changes in the configurationcan be made by altering local conditions, while selforganizationdeals with the global consequences ofthese actions (Figure 10).Figure 13. 1:1 Demonstrator at the ROM gallery Oslo / NorwayThey include move, scale of entities or change of scale,as fixing its position to force the surrounding to adapt.Color coding of elements and a navigational diagramhelps to maintain an overview of these manipulations.Precision and localization of the design model wheregiven through a millimeter based unit space and theability to link in 3D models of the site (Figure 11).paper session | Generated Lamella

346ACADIA2010life in:formation8 ImplementationThe interface allows the model to interact dynamicallywith and inside an environment given by site,program, production and material. Changes to theenvironment through manipulation are instantlyanswered by the model through topological change.These changes appear to the designer as a result ofan internal reflection rather than direct answer. In thisway designing starts by learning about the distinctcharacter of the model and its behavior.The model exchanges through customized interfaces withdifferent specialized tools: for structural FE-Analysiswith Sofistik or for the generation of production datato parametric software. The output can be adjusted todifferent model scales ranging from design speculationto 1:1 realization through machine code for Hundeggerwood joinery machines (Figure 12, 13). Intensecommunication and testing through prototypes werecrucial to determine the adequate types and dimensionsof joints, fasteners, bearing and bracing as fabricationand assembly strategies.Feedback was integrated into the model which wasbecoming noticeably aware of its placement in thebuilding process—its environment. The incominginformation was handled in a pragmatic way where newinsights were either encoded as internal conditions inthe generative code or the visual interface was used forconstraining the self-organization system.The intense preparation allowed us to exploit thecapacity of digital fabrication and self registeringjoinery, demonstrated by only 3,2 hours of cutting timeand 2 days of overall assembly.9 ConclusionSelf-organisation is a valid approach in order to designwithin complex systems characterized by high degree ofinterdependencies on the same level of hierarchy. Thehybridization of generative processes and interactivemodelling proposes shows that non-linear systems canbe used as a design tool.Here the different modelling methods are not mutuallyexclusive but work in parallel rather than in succession.Where computation is able to structure processes andrelations that are otherwise beyond human capabilities,the real time interaction offers space for designspeculation. Various constraints can be easily appliedand their effects studied. A learning process is initiatedon the side of the designer that allows him to efficientlynegotiate design intent with the systems underlyingspecifications. The interactive self-organisation modelshows thereby an “adaptive behaviour that is directedtowards an end” as Karl Popper calls it 1994. A“satisfycing” (Simon 1996) solution is achieved in theabstract space to optimise the beams’ positions. Thisbehaviour makes the presented approach more efficientthan optimising with evolutionary models (GeneticAdaption), where large numbers of options have to becreated and later dismissed in an Darwinist act. Furtherresearch into the construction of customized userinterfaces for hybrid dynamic-interactive processesmight prove valuable for opening new territories forarchitectural designThe project shows that self-organization is capable ofnegotiating in an early architectural design context.It allows to implement global design intent as well asinformation regarding production, detail and material.The advantages of this is apparent in the speed andaccuracy by which structures could be realized in 1:1(Figure13). The open nature of the approach allowsit to “learn” and become more aware of the overallprocess requirements. Further research in methods toinclude new knowledge faster and to handle the linkedimplications is necessary. This would allow to extendedthe awareness of the model to other environments andcreate new potential when gravity and tectonic stressbecome part of the initial design processFigure 14. Detail of Lamella joint in 1:1 demonstrator in OsloGenerated Lamella | paper session

ACADIA2010life in:formation347 | 412AcknowledgementsThe Lamella system was only possible through support by HSB Systems, Hundegger GmbH, Trebyggeriet.no, Knippers and Helbig AdvancedEngineering and Prof. Christoph Gengnagel/ TU-Berlin.ReferencesAllen, JS. (1999). A Short History of ‘Lamella’ Roof Construction”. Transactions of the Newcomen Society, Vol 71, No 1Bonabeau, E., M. Dorigo, and G. Theraulaz, (1999). Swarm Intelligence, From Natural to Artificial Systems, New York: Oxford University Press.Burry, M. (2005). Between intuition and Process: Parametric Design and Rapid Prototyping, in Architecture in the Digital Age: Design andManufacturing, Kolarevic, Branko (Ed): Washington, DC: Taylor & Francis.Carranza, P.M. and P. Coates. (2000). Swarm modelling, The use of Swarm Intelligence to generate architectural form”, Proceedings of theInternational Conference on Generative ArtFlake, G. W. (1998). The Computational Beauty of Nature: Computer Explorations of fractals, chaos, complex systems, and adaption.Massachusetts: MIT Press.Hensel, M. and A. Menges. (2007). Morph-Ecologies: Towards Heterogeneous Space In Architecture Design. Kfar Sava: Aa Publications.Holzner H. (1999). Entwicklung eines Nachweisverfahrens zur Bemessung von speziellen (maschinell gefertigten) Zapfenverbindungen,Diplomthesis at Institut für Tragwerksbau - Fachgebiet Holzbau, TU Munich.Kieran, S. and J. Timberlake. (2003). Refabricating Architecture: How Manufacturing Methodologies are Poised to Transform BuildingConstruction (1 ed.). New York: McGraw-Hill Professional.Kolarevic B. (2005). Towards the performative in architecture in Performative Architecture, ed. Kolarevic, B., Malkawi, A. (Routledge Chapman& Hall, 2005) pp. 205-211Popper, Karl. (1994). ‘The myth of the framework. In defence of science and rationality’ Routledge, London.Popovich Larsen, O. (2008). Reciprocal Frame Architecture. London: Architectural Press.Rittel, H. and M. Webber. (1973). Dilemmas in a General Theory of Planning, pp 155-169, Policy Sciences, Vol. 4, Elsevier Scientific PublishingCompany, Inc., AmsterdamSchwitter, C. (2005). Engineering Complexity: Performance-based Design in use”, in “Performative Architecture – Beyond Instrumentality”edited by Kolarevic, B., Malkawi, A., M., Spon PressSchindler C. (2009). Genagelt und geschraubt. ARCH+ 193, ARCH+ Verlag, Aachen 2009, p. 35Simon, H. A. (1996). The Sciences of the Artificial, 3rd Edition (3 ed.). London: The Mit Press.Tamke M. and M. Ramsgaard Thomsen M. (2009). Implementing digital crafting:developing: it’s a SMALL world . Berlin. Proceedings of theconference “Design Modelling Symposium Berlin 2009”, pp. 321-329Tamke M., Thomsen M. and J. Riiber J. (2008). Complex Geometries in Wood, Proceedings of the international conference “Advances inArchitectural Geometry”, Technical University, Vienna.Tristan, S., S. Martin and B. Daniel. (2007). Woven Surface and Form. Architextiles (Architectural Design) (pp. 82 - 89). Chichester:Academy Press.Terzidis, K. (2006). Algorithmic Architecture. London: Architectural Press.Wilson, E.O. (1975), Sociobiology: the new synthesis. Cambridge MA: Belknap Press.paper session | Generated Lamella