PARAMETRIC DESIGN STUDIO ALGORITHMIC SKETCHBOOKNINA NOVIKOVA 2015

STUDIO AIR

A1 - WEEK 1

2

lofting sets of curves to create parametric flexible surfaces that react accordingly when the base curves are shifted or manipulated.

3d voronoi apllication and deletion of lofted cells to create new geometry - carving out of a cube around a lofted shape.

3

not 110% what happened here - i think panelling tools and lunchbox trials with the panels exceeding the size of the base geometry and causing overlapping and layering beyond the contours of the shape.

below and to the left are much more successful displays of panelling applications - triangles and hexagons. panelisation is a good way to make a shape more angular, and eases on construction if fabrication is in mind, since most materials are, well, planar.

lunchbox features - panelling - diamond shape, diamonds with a second panel within, and linear layers of strips.

4

dividing surface into grid and anchoring geometry to points; rotating said geometry through ‘rotate 3d through multiple axis. the spheres are more wholesome seeing as they have the same kind of areas morphing into one another, but the cones were fun to experiment with.

A1 - WEEK 2

reconstructing something loosely in-spired by the seed cathedral (uk at expo 2010). curves are lofted to create a shape; the surface is then divided into points. each points becomes an origin point for a line extended at a z vertice. once the endpoint are identified, they can be used to mount different geometry, such as shperes.

manipulating data structures through an attractor point changes the direction in which the ‘pins’ are pointing, creating new geometry and even abstracting from the original point a little bit.

A3 - WEEK 3

6

various other geometries being used as the endpoints of a further deconstruted reverse-engineered form. extruded curcles, piped lines, spheres the size of which overshadows the use of lines (the lines, in turn, are shortened.)

7

box morph manipulation. creating a lofted surface from base curves by interpolating lines through points, and morphing on box-like surfaces by offsetting certain points and segregating the surface into panels.

A3 - WEEK 3

8

continued variation of this. it has been discovered that once the axis of how the box surface is to be extruded are changed, the new digital material no longer remains static as the outline of our base form - it comes to life, shifting, extruding in unprec-edented directions, swirling in a repetitive pattern that was quite difficult to make sense of.

edit: end of semester

in hindsight, the parameters being changed were probably the range of how much a certain part of geometry was to protrude on a ceratin vector. once the range and domain abstracted too much from the de-fault - that is, replicating the shape exact-ly, the box morph is forced outside of this confine; however as each individual panel remains interconected, fancy geometry is created between the algorithm pushing the points of morph outwards and still being anchored to the original in some points.

9

A3 - WEEK 3

10

degree of rotation - 10 degree of rotation - 60

degree of rotation - 200 degree of rotation - 200

11

editing the seed cathedral rework to have surfaces oriented on the surface instead of pipes by extruding the lines, then setting a rotatory vector to fol-low an attractor point.

in a way, this imitates the behaviour of louvres and their reaction to sunlight and their function as shading devices.

interesting to see that there realy wasn’t a drastic difference made to the overall form by changing yhe extru-sion positioning despite drastic angle change.

degree of rotation - 90 degree of rotation - 120

degree of rotation - 340

IMAGE MAPPER

12

i’m. really not sure what happened to the image resolution here. ah yes, nothing like bad internet culture

references...

projecting spheres instead of circles - intetersing to try but

everything just became much darker

high contrast is generally the most interesting effect

did you know rimbaud once tried to walk from france to russia what a fool

much better.

13

contrary to a certain belief, flipping the matrix for the script does not result in inversion of colours.

reducing the circle size creates a very faint image

projecting spheres instead of circles - intetersing to try but

everything just became much darker

much better. circles with a smaller radius actually provide better shape outline but are very very faint

curious that the white background comes up as black here

14

WEEKLY TASK - KANGAROO MESHESscrunch

curious mesh behaviour throgh relaxa-tion after shifting anchor points

derived through shifting anchor points to form two curved lines, resulting in a vault-like behaviour of the mesh.

vault

scrunch 2

even moreso curious folding with the mesh size enlarged but the cell count decreased to form bigger panels

15

cavern

i think this was achieved through introducing the y axis of unary force and shifting the points, as well as changing a number of settings on the actual kangaroo tab

dome

maximising the amount of force applied so that a relatively round shape is mainatined. anchors on the corner points. i’m not sure what’s causing the folding at the edges.

introducing the x axis and reversing the value of force applied to a negative.

reverse

16

WEEKLY TASK - RECURSIVE AGGERATION

this felt like the most beautiful unpredictable patterning, with only one axis as the dominant one

increasing rotation angle causes the tree

17

introducing another vector to which the lines extend to, creating a three-dimensional tree.

3-d extension of a single-axis driven tree as the one across.

playing around with vector values and angles to achieve a spiralling pattern.

18

PART B - ATTEMPTS

processes to commence reverse engineering of part b - loop 3 by co-de-it. preparation of base curvature through control points and amplitude variations. scaling or offsetting the curves and pulling up through a ‘move’ comand to create a multi-layer system. lofting the curves to-gether gives a nice effect, but co-de-it uses an extrusion function.

19

20

PART B - ATTEMPTS

various behaviours through kangaroo hinding function. influencing factors included mesh sizing and point count, location of hinge points and control points as well as actual bending movement imitated by the emulator (manual control).

21

22

PART C- DESIGN ADVANCEMENT

this was the first exploration of rethinking tectonic expression and structure after the

feedback received in part b. the idea was to create strong vertical elements seen as trans-parent if observed from the river itself because they’re quite thin, introduing a support system spanning through the creek. ultimatey though, this appeared quite bulky, and such a multitude of supports isn’t very justified for a bridge that doesn’t exceed 8 metres in length.

23

to be honest i’ve even lazer cut a prototype of this particular sequence, but never got around to making it. what a waste of perfectly good ply. this particular methodology seemed too akin to plain sectioning, so there’s not much regrets that it wasn’t pursued, though. the vertical panes are a very bland tectonic element and don’t complement the strips very much; the thick long spans of material take away from the transparency effect desired.

24

PART C- DESIGN ADVANCEMENT

the original truss sequence - the very first time the inklings of the system actually used for cunstruction has been introduced. base curvaure has been extruded, and the structural ele-ments matched to

25

intricate line breakdown making y- orientated connections between sweeps of points across all curvatures. unfortunately later on the sections couldn’t actually be cre-ated like that seeing as the points were not spaced evenly and the surfaces created weren’t planar, but this is kind of an estimate of how section generation could be para-metrised and remain reseptive to the digital model.

26

PART C- DESIGN ADVANCEMENT

one of the ideas - well, arguements, almost - the two collaborators had at this point was at alternative options of how the strips should be expressed. one of the suggestions was to pipe the sine cuves as opposed to extrude them - a bent metal railing seemed easier to imagine than a more or less flat patch of steel.

27

though this variation wasn’t even marked down as relevant at this point, it’s curious how similar are cables - the final solution - to something like this.