ADSAJ

AR

CH

ITEC

TUR

E JO

UR

NA

L

Architecture Design Studio: AirSemester 1, 2012

Architecture JournalCarl Madsen 357577

01rhino3d

personalstate of the art

02grasshopper

case for innovation

03the gateway

04cut definitions

05a study in moire

06realising a model

~

RH

INO

3D

WEB

INA

RS

_01

Rhino 3d (coupled with the Grasshop-per plugin, although unused in the week 1 tutorials) uses a system of point called N.U.R.B.S. - (Non-Uniform Ra-tional Basis Spline) an alternative to the mass amounts of points that go into modelling a curve in, say, SketchUp. Based on my previous experiences with computer design software such as AutoCAD, SketchUp and the Macro-media and Adobe suites, the interface was relatively easy to comprehend and follow in the videos. Although being used to the SketchUp orbiting tool, I found myself miscliking when trying to pan and orbit in the perspective view quite often, but seems to be something that I will get used to over time.

The application seems to be very fo-cused on the technical aspects of digital 3D modelling, as seen by the simultaneous projections of the model, and how seemingly powerful the appli-cation is.

I found the webinars embedded on the LMS very easy to follow, and the addi-tion of their presenter’s Rhino files was a great help in fully understanding the different methods of creating surfaces. It really felt like a progression of learn-ing, and although a lot of the informa-tion given was very basic stuff within the program, it was still nice to watch it be done, and then do it myself.

A great aspect of the online video tutorials was the ability to really learn at my own pace; to re-watch the same section a couple times if I didn’t fully understand what was happening in the video.

Overall, I am looking forward to work-ing more with Rhino; it seems like a powerful program that I will be using a lot during the rest of my degree, and likely my career.

PER

SON

AL _01

STAT

E O

F TH

E A

RT

CIT

Y O

F A

RTS

AN

D S

CIE

NC

ES

_01

Santiago Calatrava’s City of Arts and Sciences in Valencia, Spain is not only state of the art in its sleek de-sign, but also its functions as allowed by its architecture.

The long complex, made of pure white concrete and shattered tiling, is devoted to both cultural and scientific dissemination. Classically seperate, this structured city is host to the multi-functional space of both.

Its immense size provides a landmark to the city, relating with the water in a way of emergence; that the building was always there, but simply arose at the time of construction.

The planetarium section of the com-plex, the pointed oval-shaped build-ing (left), focuses on the windows pointing towards the sky, using thin, long supports to direct the eye’s at-tention to where it is pointing, using the building’s openings to create a directon.

STAT

E O

F TH

E A

RT

HES

SIN

G S

HO

WR

OO

M

_01

The luxury automobile showroom in Utrecht, the Netherlands, designed by ONL Oosterhui-Lénárd is a perfect example of how form can be developed from a func-tion and a place using methods made available by computational architecting.

Situated along the A2 highway, the extremely sleek design of the structure reflects the speed of its location, running along it in a streamlined fashion. It shows itself as iconic to the highway, presenting the qualities of the road it travels along.

Similarly, its function as an automobile showroom has large influence on the form of the building. Much like a luxury car, the showroom looks fast, aerodynamic and powerful. It even slightly resembles a car, with the ‘open’ central cockpit represent-ing the sleek top of the expensive and modern cars it displays inside.

Using parametric scripting, ONL effectively described the building in two details informed by the 3D model: one describing the highway façade, and the other the acoustic barrier which makes up the opposite side of the structure. Reference points were extracted from the surfaces, visualizing all the elementsof the exterior in simple tables, which are then run through CNC production ma-chines to create all the parts of the building.

GR

ASS

HO

PPER

AN

INTR

OD

UC

TIO

N

_02

Grasshopper is a notably powerful plugin for the Rhino3D program intro-duced last week.

What was quickly evident when first playing with the Grasshopper interface and elements was its ability to con-stantly and fluidly change extensively complex designs; the system of nodes below represents the multiple rotations of curves around the same curve. By simply changing the sliders, the design reacts in a rpedictable way, allowing the user to develop and explore the intricacies of the design. In a tradi-tional modelling application (such as SketchUp or vanilla Rhino3D) chang-ing the width of a certain element that is connected throughout the model would require a large amount of time and patience; however with Grasshop-per, this is achieved by simply moving a slider and changing a number within the program.

The possibilities for this plugin seem vast, and will no doubt be a staple en-vironment for my future designs.

CA

SE F

OR

INN

OVA

TIO

ND

IGIT

AL

TEA

HO

USE

WO

RK

SHO

P

_02

The Digital Teahouse Workshop held at the University of Tokyo in 2010 was a com-petition between three groups to produce a Japanese style tea house using the Grasshopper plugin for Rhino3D. The entirely computationally designed projects showed the varying designs of the same structure using parametric logic and concepts, as based on the differing interpretations of the brief. The constraints of the plywood offered unique challenges in the fabrication of the designs: one team etched lines in the flat CNC-routed plywood panels in order to allow them to be able to curve with the structure in waves, which would not be possible with a flat surface. Other teams decided to produce visual curves by incorporating gradual directional changes in the structure, which they adjusted to a fine point by modify-ing parametric values in grasshopper.

As an overall, the teahouses produced made a compelling argument for the appli-cability of parametric design for both large and small installations, that an ideal result can be achieved not only faster, but to a higher precision.

CA

SE F

OR

INN

OVA

TIO

NTH

E W

ATER

CU

BE

_02

“The Water Cube is one of the well recognized buildings in China where it was pri-marilly designed for the 2008 Olympics. It was also important to consider the use after the Olympics as a national aquatics centre.The building project of this archi-tecture was important to express China’s growing international role. An international competition for the aquatics centre began in 2003 to find a design appropriate for the criteria.

The current design of The Water Cube was able to win the competition as it was the most outstanding and feasible design and came up with a great concept inspired by soap bubbles.

Even though the shapes of soap bubbles seem random, their nature always touch each other without leaving any empty spaces in between and are three dimention-ally repeatable. However, to make a feasible Water Cube, there needed to be over a hundred of different ‘bubbles’.

To create this numerous complicated shapes, parametric modelling had to be used. Its role as an ‘Olympics stadium’ and its publicity and advertisement of 2008 Olympics was not the only reason of the architecture’s discourse. The unique soap bubble structure suits well with its use as an aquatics centre and creates interesting aesthetics with its complicated patterns.” (Kim, 2012)

THE

GAT

EWAY

AN

ALY

SIN

G T

HE

BR

IEF

_03

Based on the outlining of the brief, we identified four key considerations to take into account when researching and developing our design ideas. They were the promi-nent location of the site at the entry to metropolitan Melbourne, the sculpture as an iconic feature of the area, the use of the lighting in the area to create observable patterns and the taking of an experiential approach to our structure.

The site is located on the Princes Freeway at the edge of the Wyndham urban growth boundary west of metropolitan Melbourne, and the design is to act as a gateway from the undeveloped plains of the west, into the Wyndham municipality. Based on this prominent location that acts as a changing point between the relaxed landscape to the west and the faster-paced Wyndham and Melbourne areas, the sculpture should reflect this through a combination of the sharp and soft ideas that represent the contrast.

To contribute to architectural discourse and to stand as an eye-catching visual instalment, the sculpture needs to be iconic. In the context of the brief, the iconic qualities of the structure should represent its location (as described above) as well as its surrounding environment. As the sculpture is to be situated within the area of the Princes Highway, the iconic features involved should match the features of the highway. We have determined two main characteristics of the highway as iconic: speed and direction.

In line with our previous considerations, we have decided to go against our sculp-ture being objectcentred and static, opting to adopt a more experiential approach to our thinking. An experiential structure involves or is based on experience and/or observation; our sculpture should involve the audience and not just be shown to them. As such, we have focused our process on the relation of the sculpture to its audience; how it can change relative to not only the viewer’s location, but also their personal interpretation.

Again in relation to speed and direction, the low-lying ground enabled us to consid-er the effect light has as it passes through a non-filled structure; specifically, shad-ows that would lie across the road. Based on the structure, the shadows crossing the road could invoke certain feelings in the driver at certain times of the day, as well as having potentially dynamic qualities; that is, not only does the shadow move throughout the day, but changes as well.

We began our design process by cre-ating and critiquing a matrix of compu-

tational design definitions in Rhino3D and Grasshopper, producing a wide breadth of candidates that we could explore further in hopes of finding a

base for our initial concept.

The matrix consisted of input defini-tions (how the space is arranged), as-sociative definitions (how the arrange-ment of space is modified) and output definitions (how the modified space is

represented).

CU

T D

EFIN

ITIO

NS

CR

EATI

NG

A M

ATR

IX

_04

COMPONENTS

ARBITRARY POINTS

EXTRUSION ROTATION

POIN

TAT

TRA

CTO

RC

UR

VEAT

TRA

CTO

RM

ATH

SFU

NC

TIO

N

MU

LTIP

LEM

ATH

SFU

NC

TIO

N

COMPONENTS

BOOLEAN PATTERNING

EXTRUSION ROTATION

POIN

TAT

TRA

CTO

RC

UR

VEAT

TRA

CTO

RM

ATH

SFU

NC

TIO

N

MU

LTIP

LEM

ATH

SFU

NC

TIO

N

Maths Function

EXPLICIT GRIDS Extrusion Rotation Component

Multiple Maths Function

Image Sampler

Curve Attractor

00 A R C H I T E C T U R E D E S I G N S T U D I O

SURFACE GRID

ATTRACTOR POINT + COMPONENTS CURVE ATTRACTOR + COMPONENT IMAGE SAMPLER + COMPONENT

ATTRACTOR POINT + EXTRUDE CURVE ATTRACTOR + EXTRUDE IMAGESAMPLER + EXTRUDE

ATTRACTOR POINT + ROTATION CURVE ATTRACTOR + ROTATION IMAGE SAMPLER + ROTATION

The components output produced a system of curves along another curve,

be it the same one ordifferent, scaling them based on cer-

tain associative definitions. It produced a wide array of varying

results; however they seemed to lack official clarity in their arrangement. Without an in-depth understandingof how they are produced, they risk

being inaccessible to an average audi-ence. Interesting

as they may be, these abstract quali-ties were not exactly in line with our

design focus. However, withfurther experimentation, these systems could align themselves with our focus.

A system of arbitrary points across a surface or plane was one we decided

to avoid after experimentationwith it. Not only did it contrast with the ideas of parametric modelling that we

focused on previously (byusing human decision to constrain the

design), but also did not appropriate with our consideration of the

directive nature of the highway, imme-thodical.

Two explicit grid as an input is the ar-rangement of points in a square and hexagonal fashion, andhow the associations and outputs react to this difference. When juxtaposed on top of one another,the hexagonal and square patterns create differing views as the perspec-tive on the grid changes,which relates to the experiential quali-ties we are looking to produce in our structure.

The image sampler association was the use of any image found to produce a varying result, based onthe colour or darkness of certain parts of the image. We felt similarly to the im-age sampler as we didto the input of arbitrary points; it was too constrained by human intervention. There was little orderinvolved in its representation.

A S

TUD

Y IN

MO

IRE

OB

SER

VIN

G T

HE

OPT

ION

S

_05

After our analysis on the matrix of definitions in Grasshop-per and looking into the perspective differences created by overlapping surfaces and their potential to satisfy our design intent, we deemed them very similar to a moiré pat-tern.

A moiré pattern is an interference pattern that is cre-ated when two or more grids of lines (be them curved or straight) are overlaid in a non-regular fashion. This can be achieved through different shapes of lines (for example, a grid of circles overlapping a grid of straight lines), different mesh sizes or the rotation of one of the grids. The pattern is created at the intersections of the lines, where they appear thicker due to the higher density in the area. Such intersec-tions form patterns with neighbouring intersections, creat-ing a virtual line through the grids. The moiré effect (which creates the patterns) is typically an undesired effect of digitally created or altered images, but can also be used to advantages; it can be both a positive and negative effect, crafted or unwanted.

Seemingly merely a visual phenomenon created by the jux-taposition of two patterned elements, a fair amount of math-ematics is involved (differing throughout the shapes being patterned), reflecting the iconic direction approach we are looking to achieve with our design. Although the effect is visible on two unmoving grids, its more interesting qualities come from one or both of those two grids moving. However, with spacing between the two patterns, the eye moving shifts the panels relative to the viewer, recreating the effect on a still structure. This allows for a dynamic moiré pattern to be structured into the construction, changing the pat-tern it produces as the perspective differs (through both the relative speed and location of the viewer to the sculpture), resulting in a sculpture that is fast, directive, dynamic and experiential in design.

One important factor to note is that if there is to be a space between the two grid elements that form the moiré pattern, then the perspective of the viewer could potentially change the initial pattern. That is, the lines further away from the viewer would appear smaller due to them being further away, when in reality they are actually the same size as the ones closer to the viewer. This could change the pattern away from what is desired, unless it is duly accounted for.

thick lines into thick lines

basic open/close, like shutters

same width = full open, full close

thin lines into thick lines

semi close/semi open

never fully closed, never fully open, thinner lines moving between thicker

To start our experimentations with moiré patterning, we again instituted a breadth approach by creating a matrix of patterns that showcased some basic elements and how they interact with each other through changing variables. We didn’t just look at the shapes of the interacting lines, but also the thickness, angle and how far apart they were spaced. Once again we selected certain resulting patterns that both satisfied and avoided our design intent, critiquing them for further investiga-tion.

thin spaced lines into thick lines

same as thin into thick, but with inter-mittent breaks

thin spaced lines into thin spaced lines

open/close but with a large amount of time spent open as gap is larger relative to line thickness

thick lines into thick circles

creates a repeat-ing “opening” sector pattern perpendicular to the interfering lines

thin lines into thick circles

same deal, but pattern is much less pronounced as there is always a gap between the two geom-etries

15 degree thick lines into thick circles

same as first circle pattern, but a bit slower as lines take longer to be placed in the same spot (at same horizontal speed)

45 degree thick lines into thick circles

same same but slower once again

thick circles into thick circles

creates “opening” pat-tern in centre, rever-berates around entire combined geometry

thin circles into thck circles

“opening” pattern not visible, but shapes made from the gaps are very pronounced

thin circles into thin circles

same as first but hyper mode (faster, more, etc)

thin spaced circles into thin spaced circles

same deal with lines, still there but little inter-ference

128 line radial into same

crazy intricate pat-terns all over the shop

64 line radial into same

less patterns, but clearer

32 line radial into same

again less patterns, too many to be pro-nounced it seems

thin spaced circle into 16 line radial

some patterning, trumped by the dif-ference in geometries

Our initial generalised views were that lines that were simi-lar (but not necessarily an exact) thickness and distance apart seemed to produce more pronounced patterns within themselves, whilst lines that were exceedingly different thicknesses and distances apart had patterns that were very vague and hardly noticeable; on a high-speed high-way these light patterns would be overlooked and lost.

Rotation of lines into linesThe rotation of a set of lines into a similar set of lines shows the shrinking of the moiré lines into themselves as the angle of rotation between the grids increases (or the stretching of lines as the angle decreases). Although an interesting con-cept, the rotation of one grid without the other cannot be plausibly achieved in a static structure.

Lines into circlesThe introduction of a set of lines into a set of circles cre-ates a very interesting opening/closing effect that runs horizontally against the vertical lines, matching the flow of the highway. The speed at which the moiré lines open and close is determined by the speed at which the grids pass each other (represented by the speed at which the motorist travels).

Lines into linesThe movement of lines between lines created a very sim-plistic shutter-like opening and closing, which we deter-mined to although have relevance to the vectorial (cre-ated by speed and direction) qualities we were looking to achieve, its pattern was too basic and uninteresting to be developed beyond that.

Radials in generalLines arranged in a radial fashion superimposing on each other produced very intricate and omnidirectional curves in their boundaries, but this effect was overshadowed by the extreme clutter that developed at their centres. This clutter looks out of place in relation to the cleanliness of the sur-rounding ellipses to the point where it almost overshadows the moiré pattern created, resulting in a very undesirable connection.

A S

TUD

Y IN

MO

IRE

DEF

ININ

G A

TYP

E

_05

Based on the horizontal directional and easily distinguishable patterning qualities of the moiré effect created by the juxtaposition of a group of circles of varying radii with a grid of lines moving into them, we decided to have a more in-depth look at why this particular effect is created, and what the altering of variables within it pro-duces, finding some very interesting.

Based on a pre-existing paper Research on the Moiré fringes formed by circular and linear grating (Chen et al 2011), we discerned that the spacing of the lines played an important role in the moiré patterning visible between circles and lines. A succinct summary of the paper is the focus on two variables, P and a, where P is the distance between the circles and a is the distance between the parallel lines (summarised visually below).

Three distinct shapes were realised when changing the variable P (the distance between the parallel lines) against a static a. When the ratio P/a was greater than 1 (that is, when P > a), ellipses formed horizontally along the grids. When the ratio P/a was equal to 1 (that is, when P = a), parabolas formed, focusing to the centre of the circle. When the ratio P/a was less than 1 (that is, when P < a), hyperbolas formed, focusing to the centre of the circle. However, when the ratio P/a became too far away from 1, the patterns became difficult to distinguish, resulting in the need to care for the distances. This radical change in moiré patterning based solely on the distance between parallel lines prompted us to develop a parametric defini-tion in Rhino3D and Grasshopper that would allow us greater control over the vari-ables involved and the outcomes produced.

The definition we created in Grasshopper had six useful variables: P, the distance between the centres of the parallel lines, a, the distance between the centres of the circular lines, the thickness of the parallel lines, the thickness of the circular lines, the number of parallel lines and the number of circular lines. This enabled us to quickly test a large number of various iterations of the created patterns in real time by merely moving a slider.

On reflection of the potential issue (brought up in the introduction to the moiré ef-fect) that due to the viewer’s distance from each grid of lines they would see a dif-ferent pattern than what was intended on a flat surface, our experimentation with the moiré grasshopper definition presented an exceptionally exciting solution. We noticed that as the parallel lines became closer together (that is, as P decreased); the moiré pattern would quickly turn from ellipses into parabolas. However, with calculated spacing between the parallel lines and the circular lines, the viewer’s perspective of how many parallel lines were attributed to how many circular lines would be relative to their position. Somebody standing on the side of the parallel lines would see many circular lines for every parallel line, as the circular lines would be further away, appearing smaller to the viewer in relation to the parallel lines. However, somebody standing on the side of the circular lines would see the op-posite; more parallel lines for every circular line. With sensitive variables such as P and a, this effect compounds into a viewer on the circular side seeing hyperbolas, whilst at the same time a viewer on the parallel line side seeing ellipses. The way these two moiré patterns change as the viewer moves across horizontally relates perfectly into changing views depending on whether or not the sculpture is viewed coming into the city, or whether they are leaving the city. The sharpness of the mov-ing hyperbolas would relate the driver coming into the fast-paced nature of Mel-bourne city, whereas opening ellipses would give a softer feeling, ideal for some-body travelling away from the commotion of the city.

To visualise this idea, we rendered perspective views of the two different sides of the concept in Rhino, showing the dramatic change that occurs just by being on the other side of the sculpture.

REA

LISI

NG

A M

OD

ELB

AN

Q R

ESTA

UR

AN

T

_06

“The design intent of BANQ Restaurant shows layers of contour lines on top of each other and this to shows curving shapes that resemble waves or carved out sedi-ments affected by the movement of water. The form itself creates calming sensation in the environment and also the changes in the shapes and physical form shows interesting silhouette. This silhouette can be affected by different directions and strengths of light creating soft and beautiful shades that shows strong silhouette depending on the lighting source.

We want to focus on it’s movement and changes in form and how it affects the shadows and silhouette with changes in the light source.

We were given definitions to experiment on with parametrics and to explore and fo-cus on the ideas, such as movement, silhouette and speed, we reverse engineered BANQ Restaurant (as it had similar design intent) to generate and extend our ideas.

Data Driven Component was the major definition that reflected our ideas of various curve lines. From these lines created, we made surfaces and subtracted surfaces in between the lines to create a form - which then repulicated and layed them on top of each other to see the form it creates.

We needed to experiment with the lighting and the silhouette, so a physical model had to be made. The overlapping of the shadows of the physical model creates interesting silhouette. If cars were to drive through it, it could create flickering of lights, increasing sense of speed and movement.

This process was taken to generate ideas and this was not enough to create con-tour lines that we intended, so we moved on to creating different Grasshopper defi-nitions.” (Kim, 2012)

“Using the Moire patterning and the BANQ Restaurant contours as concept, we were able to combine these two ideas together to generate this parametric model using the definition above.

The steps taken:1. Two thin boxes were drawn on Rhino layered next to each other.2. The front facade will show the contour lines inspired by BANQ Restaurant. By using the commands ‘Rebuild’ and ‘PointsOn’, the surface of the front box became free to change the form to create the contour.3. We used the above definitions to create linear Y-Z contour lines on the boxes.” (Kim, 2012)

placeholder