Jonathon Belotti (539430)Semester 1 2013

AIrArchitecturedesign Studio

Design Approach

Contents

Design Focus- Structure1

STRUCTUREdesign Focus

It is not necessarily explicit, but al-ways implicit in the design of buildings. Structure, as it relates to architecture and construction, is a body or a system of bodies capable of supporting loads, and resolving them into a stable equi-librium.1 It is a constant restriction on architecture, governing material use, form, and scale. Though it is conceiv-able, a paper thin layer of concrete cannot serve as a roof structure, and it thus architectural unfeasible despite its aesthetic utility in architecture form. The Wyndam Gateway Project, it is ar-gued, ought to be an installation draw-ing from the theory of parametric and computational architecture, and de-signed by its methods, technologies and processes. As a computationally design installation, there is potential for complexity in the composition, in-terrelation and organisation of its ele-ments. This is exemplified by the strik-ing Metropol Parasol in Seville, Spain. Designed by Jürgen Mayer-Hermann, its form is intricate and complex, its untested and experimental laminated wooden structure encountering struc-tural issues that increased costs. The instability of the structure was eventual-ly corrected by application of high-per-forming glue though this precedent de-sign still demonstrates the importance

of good structural design as a com-plement to good parametric design.2

Good Structural Design will enable the physical realisation of an ambi-tious, intricate and mathematically complex installation, in general effort directed towards the solving of struc-tural problems frees up the design to explore further computer-aided design potentials. On a more som-bre economic note, consideration of structure can save money in construc-tion and material costs. Norman Fos-ter’s Heart Tower utilised a diagonal-ly gridded steel structural (a Diagrid), which required 21% less steel tonnage than a conventional steel structure.3

Structure in parametric and computa-tional discourse can be identified as containing a few current activities. The first is the exploration and experimen-tation with the grid as an organisation-al structure, translatable into physical structure. The orthogonal grid has long been utilised to organisation building elements for the withstanding of loads, and it reflects closely our abstracted understanding of gravitational and lat-eral load on structure. However, the or-thogonal grid is just one type of the lat-tice algebraic structure.4 In modernistic architecture, rectilinear 3 dimensional

Metropol ParasolARCHITECTS: jurgen mayer hLocation: Seville, spain2005-11

Hearst TowerARCHITECTS: Norman Foster Location: New York, USA2006 metropol parasol

lattices could be easily related to the rectangular mass employed predomi-nantly by architects. However, within a parametric landscape, constantly pro-ducing curvilinear and blob-like vol-umes (or ‘soft’ form)5, the rectangular grid is less relatable and useful to orga-nise building elements into a structure. Mathematically complex, though visu-ally intuitive, lattices can be created that conform to the new forms and volumes of computational architecture. Currently the omnipresent issues of

loading and force equilibrium are be-ing resolved in structure in more com-plex ways, and the physical eventual-ity of these solutions is the lattice of structural elements. Examples of this can be seen in John McAslan’s Kings Cross Station Concourse and the Ja-pan Pavilion by Shigeru Ban. Both use a lattice of structure members (steel and bamboo respectively) to span a space. The evolution in structure is striking when these works of archi-tecture are considered against post and lintel, truss, or arch structures.

KING Cross Station consourseARCHITECTS: John mcaslanLocation: london, uk2012

Japan PavilionARCHITECTS: Shigeru BanLocation: hannover, Germany2000

KING Cross Station consourse

A second activity of Architects and en-gineers related directly to structure is the linking of structure to parametric geometry.6 Structure is effectively un-derstood visually, showing elements and how they are being influenced by forces. Graphic statics are a collec-tion of geometric algorithms based on vector calculus that aide in the find-ing of equilibrium solutions, where a form is stable and at rest.7 In compu-tation, work has been done to enable programs to better link the geomet-ric design process to the structural analysis process of Graphic Statics.

There has been success and currently parametric models of geometry can be analysed to inform designs to a de-signs structural feasibility, especially in the early stages of parametric design. 8Karamba is a finite element program

for Grasshopper 3d that does exactly this, enabling designs to test structural feasibility during form-finding process-es, and also structural optimisation af-ter form has been finalised.9 Being a finite element program, structure is an-alysed by inputting material and struc-tural properties into a mesh, which through algorithmic calculation yields the expected response of the geometry to tensile and compressive loading.10

Currently, designers can use a program to have an immediate and dynamic connection between the geometry, its structure and themselves.11 Because of research and development into the linking of structural analysis to para-metric geometric models within the design workspace, architects increas-ingly have an amplified ability to create form that is considerate of its structure.

Screencapture of grasshopper with karamba plugin modules utilised to inform, in real time, the designer of the effects of loading on parametric geometry

Screencapture of a grasshopper geometry that is being visually informed by structural analysis. The use of specific modules allows the geometric elements to represent, force magnitude, force type and force direction as well as remain ing a dynamically modifiable geometric form

Finally, the exploration of form in the parametric/computational landscape has been aided by greater understand-ing of deeper biological, chemical and atomic structure in science. Contempo-rary architectural practice is increased by observed to be expressive of these unseen structures in nature, whether or not the mathematical significance of those structures has been utilised in the building. The Harpa Concert Cen-tre by Henning Larsen Architects is one such building that is expressive of structure that is not necessarily an em-bodiment of its own structure.12 Its form was partly inspired by the structure of

basalt crystals (see right) However, the Water Cube by PTW Architects, while being inspired by soap bubbles cellu-lar structure that forms the geometric organisation of its steel beam skeleton, and the result was a highly stiffened building capable of remaining static if turned on its side.13 Aranda\Lasch Ar-chitecture also, quite uniquely, have utilised the structure of quasi-crystals to produce variation and infinite combi-nation in form from a small set of re-peated elements. The result is a struc-tural entity of similar structural quality to a rectangular grid structure, though with differences in form and possibility.14

harpa concert centreARCHITECTS: Henning larsenLocation: reykjavik, iceland2011

Beijing Aquatic centreARCHITECTS: PTW architects Location: beijing, china2008

Beijing Aquatic centre

quasi-crystal structuresarchitects: aranda\lasch

1 Pullan, Wendy, Structure (Cambridge: Cambridge University Press, 2000)2 Carrasco Purull,Gonzalo & Livni, Pedro, Paranormal Activity: Metropol Para-sol and other Para-tech. (2011) <http://vostokproject.com/2011/05/27/paranor-mal-activity-metropol-parasol-and-other-para-tech/> [accessed 8 April 2013]3 Sveiven ,Megan, Flashback: Hearst Tower / Foster and Partners (2012) <http://www.archdaily.com/204701/flashback-hearst-tower-foster-and-partners/> [ac-cessed 8 April 2013].4 , Lattice (order) (2013) <http://en.wikipedia.org/wiki/Lattice_(order)> [accessed 8 April 2013].5. Schumacher, Patrik, Patrik Schumacher on parametricism- ‘Let the style wars begin’ (2010) <http://www.architectsjournal.co.uk/the-critics/patrik-schumacher-on-parametricism-let-the-style-wars-begin/5217211.article> [accessed 27 March 2013].6 Preisinger, Clemens. "Linking Structure and Parametric Geometry." Architec-tural Design 83, no. 2 (March 2013): 1117 Lachauer, L., Junhjohann, H., Kotnik, T.: Interactive Parametric Tools for Struc-tural Design. In: Proc IABSE-IASS (2011)8 LACHAUER L., KOTNIK T., “Geometry of Structural Form”, in CECCATO C., HESSELGREN L., PAULY M., POTTMANN H. and WALLNER J. (eds.), Advances in Architectural Geometry 2010, Springer, Heidelberg, pp. 2019 Preisinger, Clemens. "Linking Structure and Parametric Geometry”, 11210 Widas, Peter, Introduction to Finite Element Analysis (97) <http://www.sv.vt.edu/classes/MSE2094_NoteBook/97ClassProj/num/widas/history.html> [ac-cessed 8 April 2013].11 LACHAUER L., KOTNIK T., “Geometry of Structural Form”, 20112 Henning Larsen Architects (2011) <http://www.dezeen.com/2011/08/25/harpa-concert-and-conference-centre-reykjavik-by-henning-larsen-architects/> [ac-cessed 8 April 2013].13 ‘Bubble Engineering’, Catalyst, ABS Television, 26 March 2009. <http://www.abc.net.au/catalyst/stories/2526116.htm>14 Burrichter, Felix, Document No. 158—In the studio with Aranda\Lasch (2012) <http://documentjournal.com/document-no-158/> [accessed 8 April 2013].

Footnotes

STRUCTURECASE STUDY 1.0

1.- A simple loft

2.- Using LunchBox we used the command random quad panels

3.- After we used the random split command and sep-arated the quad panels into two different surfaces. We then used an offset to move some of the panels away from others?

4.- Applying the same concept on a torus surface and changing the panels to diamond panels we get the next outcome. The beginning definition is the next one:

5.- We then created a plane that intersected the torus through the middle and used a bolean operation to cut our original form in half.

The use of the plugin LunchBox for grasshopper is of great help to create forms more attached to reality. Not only because the structures are used frequently but also because it has been proven they work. In this exercise we attempted on creating a structure for a promenade, some of the panels were randomly deleted allowing natural light and ventilation into the walk. The center could be used as a garden with different kinds of plants and some benches for the people who are tired of walking to go rest.

We later applied diamond structure to the surface and with number sliders we were able to change the num-ber of diamonds we got on the surface

Experimenting a little we found out how the geometry of the surface changed according to the changes we made with the number sliders to the U and V division parameters.

We later applied diamond structure to the surface and with number sliders we were able to change the num-ber of diamonds we got on the surface

Experimenting a little we found out how the geometry of the surface changed according to the changes we made with the number sliders to the U and V division parameters.

STRUCTURECASE STUDY 2.0

king’s Cross Concourse

Prototypes

the abandoned Tensegrity model This model was developed as a repeatable stackable structural module, much resembling the structural arrangement of the Beijing Water Cube.

This model however is the beginnings of a geometry that acheives stability with tensegrity. However, the design turned out to not a viable within the Grasshop-per design space.

PLate structure modelMaterials used in modelling the Grasshopper sketches should, I beleive only neccessar-ily reflect the structural qualities of the true material. This also, need only be taken as far as neccessary for modelling, as in, wood is a n adequate representative material for steel, as small scales, although not being structur-ally equivical.

The phenomenalogical features of the mate-rial, its touch, appearance, sound etc need only be bonuses of representation. The importance of any model is in its ability to demonstrate the structural logic.

How elements of a model fit together, and their orientation is a big issue to overcome. This paper model is crudely and unreal-istically bonded with adhesive. The use of adhesive structural bonding is not unheard of, the Metropol Parasol in space required it. However this fabrication of a portion of an algorithmic sketch uses adhesive inappropi-ately.

Card is not an ideal construction materal, platics or wood could serve better.

TECHNIQUE PROPOSALCurrently of grasshopper techniques do not respond to topography. The forms are quite easily manipulated to be situated comfortably within the site boundar-ies though it is certain that there in not an intelligent relationship between the structural algorithms in Grasshopper 3D and the site.

However the techniques we have employed that are based around the use of algorithm modules that can develop structures that within a computational design space. That itself is quite groundbreaking however the design also have shown they have potential to be highly efficient, flexible and adaptable to the needs of the client.

Form follows structureThese linework explorations were created around the idea creating a decora-tive cladding, to alter the structures form, that would be developed directly from the formal characteristics of the structure.

Although it is visually interesting to see the design self-referencing its own structure within its form, I was decided that this type of decorative cladding was ad-hoc and unneccessary.

Learning objectivesand outcomes

Earlier readings on the discourse sur-rounding the emerging parametric and computational theory and prac-tices within architecture, greatly aid-ed in investigations of the Gateway brief. Reading about recent and cur-rent work and research utilising the capabilities computational tools and programs, I felt confident shaping a design approach that could sensi-tively respond to the brief, the chal-lenges of design and the limitations of computational tools and programs. Choosing to explore the changes in structure in Architecture, the study of relevant precedents helped to form convictions as to the proper direc-tion for a parametric exploration. Most powerful were the ideas that compu-tational design, with its complex math-ematics, could free structure from its long-serving partner, the orthogonal grid. The Hearst Tower by Norman Foster employed recently a dia-grid in place of an orthogonal, rectilinear, and traditional lattice to arrange its struc-tural steel members. The result was more efficient, utilising less material. I was heartened, believing that new ap-proaches would not just bring differ-ent structures, but better structures. It seemed that the new relation-ship between computer and designer could provide a strong rational for a

design, solving an important issue in the creation of architecture, stability. Unfortunately what followed was a confidence shaking, productivity de-stroying lesson in the difficulties of designing within a computational de-sign space. It was an objective to “generate a variety of design possi-bilities”, instead I feel our efforts pro-duced little possibility and instead a variety of different broken algorithms. LunchBox was suggested as a starting point for exploring structure in Grass-hopper, but as a starting point it was facile. Everything happened so easily, and this seemed to be because Lunch-Box could neatly perform the limited tasked it was designed to, with little input, thinking or understanding from we, the designers. Having recreated the form of the Hearst Tower struc-ture and the Kings Cross Concourse, I gained no real understanding of why these structures were stable or useful. Recognising the limitations of the LunchBox plug-ins, the group made forays into the Kangaroo plug-in, but with what could not be described as success. A lack of understand-ing of both the programmes work-ings and the deeper algorithmic workings turned out more ‘broken algorithms’. This disconnection be-tween the designer’s intentionality and

the creative process was a new and difficult problem to deal with. This was in part due to structural analysis being a foreign skill to the architecture stu-dent (engineers business), and also the added difficulty of tackling with this skill within the computational domain. These things compounded the diffi-culty of the task, which led to a certain feeling for me, of being lost at sea.What has eventuated was a somewhat myopic and tunnel vision ap-proach to our design. We wanted the Kangaroo, and subsequent Karam-ba, plug-ins to function! This became also our entire goal, we needed to demonstrate the ability of Grasshopper 3D as a computational design space to deal with the architectural problem of achieving structural stabil-ity and efficiency for the parametric ‘soft’ and ‘complex’ forms being produced.Stepping back from working in Grasshopper, I believe I regain the clarity to “make a case for proposals”. Even with an admittedly limited reading through of contem-porary architectural discourse, an argument has formed that I believe can combat accusations of superficiality, vacuity and purposelessness aimed at the developing parametric designers. We have found that when it ‘clicks’, designers can have a pow-erful immediate interaction with structure as it relates to parametric geometries.