Digital Design - Module 02 Semester 1, 2018 Francis Burne Thompson (757 758) Samuel Lalo Studio 11

Week Three

Reading: Kolerevic B. 2003. Architecture in the Digital Age

Kolerevic described three fundamental type of fabrication techniques in the reading. Outline the three techniques and discuss the potential of Computer Numeric Controlled fabrication with parametric modelling.

Digital fabrication techniques can each be sorted into one of three categories: Subtractive, additive or formative. Subtractive techniques begin with planes or volumes of material that are then cut or otherwise machined to remove excess material, leaving the desired part or object. This results in a fabricated part with no weakening joins but results in a large amount of wasted material and certain geometries (such as overhangs) can only be manufactured by more advanced (and therefore expensive) 4 or more axis milling machines. Examples include laser cutting and CNC milling. Additive techniques work by building up layers of material. These processes produce minimal waste but can be fairly slow and produce fragile parts of limited dimensions, although this does vary. 3D printing in all of its forms are the most common examples of additive manufacturing. Formative manufacturing involves the reshaping or moulding of various materials such as piping or laminated timber. Material is reshaped with carefully applied mechanical forces and may involve treatment beforehand to alter the physical properties of the material. Computer numeric controlled fabrication allows parametric modeled designs to be built by allowing the highly varied forms and components to be constructed with minimal cost additions compared to manufacturing an equal number of identical components.

2

Week Three

Surface Creation

The definition for surface creation is made of repeating units. 8 points create 4 lines which culminate in 2 surfaces. While iterating through surface variations I quickly decided on a roughly pyramidal form. The iterations therefore vary the central void between the two surfaces. Iterations one and two feature one touching corner to create a wedge shaped void. This was changed to a much larger void in iteration 3, however the curvature of the surfaces was a) too symmetrical to differentiate each surface b) limiting for panel variations. The final iteration has no touching points between surfaces to create a angled rectilinear void and two distinct surfaces.

3

Week Four Panels & Waffle

Both surfaces were panelled with the same repeating units, a combination of a 2D panel and a 3D volume, each of which appearing in a large size and small size. The small size has a base with half the dimensions of the large but the height remains unchanged. The panels are interleaved to transition from large weighted volumes at the base with clusters of smaller volumes finally falling away to leave only large panels. This patternation is intended to imitate the barnacled forms of protruding ocean rocks while allowing light to penetrate the entire structure.

4

The horizontal elements of the waffle have been reduced to remove the upper most fin, creating a central light well. The vertical members have been aligned perpendicular to the diagonal between the two surfaces, which when combined with the offset internal edge, lift the entire construction and create openings which perforate the base edge.

Week Four

Laser Cutting X1 9

The laser cut file creation was an exercise in compromise between cost (efficiency) and buildability.

X1 13

The waffle fins (left) are individually distinct and thus easy to distinguish in the final product. As a result efficiency could be maximised and the buildability could be sacrificed by removing labels and sharing edges.

X1 12 X1 10

X1 11

X1 8

In contrast the webs of the paneling (right)are not easily identified. This meant that nesting with shared edges was avoided and labels added in order to easily build the surface panels. Etch lines were also used to avoid

X1 6

X1 7 X1 3

X1 4

X1 1 X1 2

X2 11

the use of masking tape, which has a tendancy to damage the more delicate ivory card.

X1 5

X2 10

X2 12

X2 6

X2 4

X2 2

X2 8

X2 3

X2 5 X2 1 5

X2 7

X2 9

Week Five

The creation and transformation of the boolean shapes was done entirely in grasshopper, with no rhino geometry referenced in. Because the geometry was created with a grasshopper definition the length, number of sides and tapering could all be changed fluidly (top left). This shape was then stretched and rotated, again done in grasshopper to easily alter and iterate the shape (centre left). The volumes were finally scaled according to the size of the grid cell they occupied - the scaling ratio could be altered fluidly with number sliders (bottom left).

6

Week Five

Isometric

This arrangement and configuration of voids was chosen because of the way in which the voids intersect produced interesting results. Horizontally the voids align and create stretches of space while vertically the voids punctuate each other, but they do not create long stretches of openings. Where the voids interact with the edge of the cube volume the honey comb like nature of the shapes and their tesselation is articulated. This section was chosen at it demonstrates the honeycombesque openings on the rear face of the volume as well as sectioned cuts along the long axes of the hexagonal voids. The size of the voids was positively correlated with proximity to the base of the cube. This allows the voids to combine towards the base and form larger volumes while smaller volumes perforate the ceiling of the larger volumes as well as the edges of the cube. This creates secluded or private areas the branch off from the larger main volume. Despite the complex network of voids only a single, albeit large, opening exists at the base. This means that this would not be a space to move through but instead a space to pause or stop within before moving on.

7

Base Surface

{0, 75, 150} {150, 37.5, 150} {150, 75, 150}

{150, 37.5, 150} {150, 75, 150} {0, 0, 0}

{37.5, 150, 0}

{0, 150, 0} {112.5, 150, 0} {150, 150, 0}

{150, 150, 0}

{150, 150, 0}

1.1

{150, 0, 0}

{112.5, 0, 0}

{37.5, 150, 0}

{37.5, 150, 0}

{0, 0, 0} {0, 37.5, 0}

{0, 0, 0} {37.5, 0, 0}

{112.5, 0, 0}

{112.5, 0, 0} {150, 0, 0}

{150, 37.5, 150} {150, 75, 150}

{150, 37.5, 150}

{0, 0, 0} {37.5, 0, 0}

{150, 150, 0}

1.2

1.3

1.4

Grid Point Distribution

{170, -6, 132}

{112,142,25}

{112,45,-30}

U/V = 10 2.1

2.3

2.4

3.2

3.3

3.4

Panel Geometry

2.2

3.1

Iteration Matrix 1:5 @ A1

Task 01 Matrix Task 1 was an exploration in the creation of a gradient from closed to open forms. With this in mind the surface iterations gradually increased the upper opening until a comfortable amount was found that did not overly curve the surfaces. The second layer of iterating was based around the altering the visual weight of the panels with attractor points. The large amount of U/V divisions iterations was chosen with intent to interleave a number of panels of different sizes to achieve this effect. With this in mind all 4 of the panel variations were chosen for the final product, with the panels moving from 3D to 2D with smaller panels to mediate the transition

8

Grid Transformation Random 1.1

1.3

1.4

2.2

2.3

2.4

3.2

3.3

3.4

Void Shape

1.2

2.1

Void Orientation and Scaling 3.1

Iteration Matrix 1:5 @ A1

Task 02 Matrix The intent behind this model was to create a central void band or cluster without removing too much material and maintaining the edges of the individual volumes. To this end the grids were transformed with attractor curves that intersected the entire cube. The final iteration was chosen because it achieved the desired effect without creating unnecessary noise. The boolean volumes were created in grasshopper which allowed rapid iterating on the shape itself. An elongated, tapered hexagonal shape was chosen because tessellate well without removing too much material. The orientation of this shape was chosen as a diagonal because purely horizontal or vertical orientations intersected too little and too much respectively.

9

Week Six

Final Isometric Views

10

Appendix

Process

The

required

The orientation of the 3D printed form

several iterations in order to maximise build time without passing the 9 hour limit.

3D

printed

section

allowed near zero support to be included

11

Appendix Process

Documented strange offset behaviour of open curves in grasshopper required curves to be offset in both directions, these sets of curves were then sorted to find the internal offset by culling alternating list items before finally lofting the resulting curves with the original contours to create the vertical waffle elements

12

Appendix

Process

(Left, (Right)

centre) Opposing views of each surface. Photo of surface not visible in main photo

13

Fburne 757 758 module 02 journal
Fburne 757 758 module 02 journal