Presentation on theme: "ISAMA 2004 Artist’s Sketch, SIGGRAPH 2006, Boston, Carlo H. Séquin, EECS, U.C. Berkeley Ling Xiao is an undergraduate student who worked with me on."— Presentation transcript:
1ISAMA 2004Artist’s Sketch, SIGGRAPH 2006, Boston, Carlo H. Séquin, EECS, U.C. BerkeleyLing Xiao is an undergraduate student who worked with me on this project for the last 12 months. He did all the programming and fancy rendering for this talk-- while at the same time he also worked on 3 other projects for 3 other professors. Unfortunately he cannot be here today because he is busy finishing up some of these other projects, preparing himself fro graduate school at Stanford, and starting a new company.The title is somewhat cryptic – but to first order this is about …Hilbert Cube 512
2a “space-filling” curve 3D Hilbert Cubea “space-filling” curve
3A plane-filling Peano curve The 2D Hilbert Curve (1891)A plane-filling Peano curveFall 1983: CS Graduate Course: “Creative Geometric Modeling” Do This In 3 D !
4Artist’s Use of the Hilbert Curve ISAMA 2004Artist’s Use of the Hilbert Curveone conceivable answer might be this artwork by H.F. ... but still a 2D pattern, just on a more complicated 3D surface.Helaman Ferguson, “Umbilic Torus NC” Silicon bronze, 27 x 27 x 9 in., SIGGRAPH’86
5Construction of the 2D Hilbert Curve ISAMA 2004Construction of the 2D Hilbert Curve123To understand how we might truly carry this idea to 3D, take another look at the construction ...
6“Do This in 3 D !” What are the plausible constraints ? 3D array of 2n x 2n x 2n verticesVisit all vertices exactly onceOnly nearest-neighbor connectionsFill “local” neighborhood firstAim for self-similarityRecursive formulation (for arbitrary n)
8Construction of 3D Hilbert Curve Use this element with proper orientation, mirroring.
9Design Choices: 3D Hilbert Curve ISAMA 2004Design Choices: 3D Hilbert CurveWhat are the things one might optimize ?Maximal symmetryOverall closed loopNo consecutive collinear segmentsNo (3 or 4 ?) coplanar segment sequenceothers ... ? More than one acceptable solution !Is the answer no fully predetermined, or is there some freedom of design ?
10Typical Early Student Solution Design Flaws:2 collinear segmentsless than maximal symmetry4 coplanar segmentsD. Garcia, and T. Eladi (1994)
11Jane Yen: “Hilbert Radiator Pipe” (2000) Flaws ( from a sculptor’s point of view ):4 coplanar segmentsNot a closed loopBroken symmetry
12ISAMA 2004Time-Line, BackgroundDavid Hilbert, Construction of a 2D Peano curve (1891).E. N. Gilbert, “Gray codes and the Paths on the N-Cube” Bell Syst. Tech. J. 37 (1958).William J. Gilbert, “A Cube-filling Hilbert Curve” Mathematical Intelligencer 6(3) (1984).C. H. Séquin, “Do This in 3D!” Graduate course assignments ( now).Nelson Max, “Visualizing Hilbert Curves” (VIS’98); “Homage to Hilbert” computer-generated video.C. H. Séquin, Plastic models (1998). C. H. Séquin, Metal Sculpture (2005). Have these students invented something completely new ?1958 BSTJ Coding theory !
13Plastic Model (from FDM) (1998) Support removal can be tedious, difficult !
15The Next Level of Recursion Presented a challenge to remove supports.Resulted in a flimsy, spongy model.Would like to have a more durable model in metal.
162006: Metal Sculpture in Exhibit ISAMA 20042006: Metal Sculpture in ExhibitFinally got around to doing a metal sculpture last year using a special layered sintering processDesign:closed loopmaximal symmetryat most 3 coplanar segments
17The Devil is in the Details ! ISAMA 2004The Devil is in the Details !Aesthetic design goals dominated.Abandoned strict self-similar recursion.Used a different lowest-level unit element.Moved top-level connections to center.Strict S4-symmetry could be obtained.This solution could not have been found without computer-aided design tools.To achieve the various aesthetic design goals, and some constraints imposed by the fabrication, strict self-similar recursion had to be abandoned.To avoid 4 coplanar elements, the 8-segment Hilbert path, at the smallest unit level, was broken at one of the 4 mutually parallel segments.To move the connections at the largest scaleto the center of the sculpture (needed for support; more on that later ...),half the smallest unit cubes were rotated 90°.Overall strict S4-symmetry could be obtained.
18Basic Element, Lowest Level not this – but thisavoid 4 coplanar segments !
19Implementation Challenges How to build this in metal ?Impossible to get machine tool to inside;Hard to cast; complex mold;Fortunately, new process from X1 corp.
20New Metal Sintering Process ProMetal is a division of The Ex One Company headquartered in Irwin, Pennsylvania USA.Ex One, known for innovative technologies, incorporates the ProMetal process to their line of products and services providing an advanced manufacturing solution.
21PROMETAL Printing Process Selectively, layer by layer, infiltrate metal powder with a binder (like “3D printing”).Remove all un-bound metal powder.Sinter the remaining “green” part; stainless steel particles fuse, binder gets flushed out (hopefully in that order!); porous (50%) stainless steel skeleton.Infiltrate with liquid bronze alloy; fully-dense composite.
22Problems ... Green part is heavy, but not very strong. My sculpture is a 320” inch long rod, 3/16th” thick, wound up in 4” cube, with no intermediate supports.Green part needs additional supports !! We started with 12, but needed 36.Finally these supports need to be removed again; put them near periphery for easy access.But center also needs some supports (which would be hard to cut away); make these the permanent ones.This necessitated one more redesign ...
23Auxiliary Supports for “Green” Part ISAMA 2004Auxiliary Supports for “Green” Partfat green tubes near center are the permanent connections between the two halves of this “cubist brain”
24The Two Halves of the “Cubist Brain” View along a symmetry axis
25Of Interest to Siggraph Attendees: ISAMA 2004Of Interest to Siggraph Attendees:New fabrication process: allows to build things not previously possible.Show the intricate design challenges behind a relatively simple sculpture.What are its artistic merits ? What associations does it raise ? . . .Give you a glimpse of my creative process: Open-ended analogies intriguing results. Another example: 3D Yin Yang. Why do I think this sketch may be interesting to SIGGRAPH attendees ?. . .What are its artistic merits ? I find it an intriguing and rather pleasing form. Ideally good art speaks for itself and does not need lengthy explanations. Associations: A cubistic constructivist picture of the human brain ... Two halves loosely connected ...
26Design Problem: 3D Yin-Yang “Do this in 3D !”What this might mean ...Subdivide a sphere into two halves.
273D Yin-Yang Solutions (Fall 1997) ISAMA 20043D Yin-Yang Solutions (Fall 1997)Amy Hsu: Clay ModelPerhaps the most obvious solution ...Robert Hillaire: Acrylite Modeland these students are in good company ...
28Max Bill’s “Half-Sphere” Max Bill, Swiss ( )“Hard Half of a Sphere”Fused silica, 18 in. diameter(1972).
36Collaboration with Brent Collins ISAMA 2004Collaboration with Brent CollinsBrent Collins is a professional artist living in Gower, MO,who has been carving abstract geometrical structures fromsolid wood blocks or from laminated assemblies. Many of hissculptures comprise minimal surfaces which form an intricatecomposition of tunnels and saddles.“Genesis” – Brent Collins at BRIDGES 2000
37“Hyperbolic Hexagon” by B. Collins 6 saddles in a ring6 holes passing through symmetry plane at ±45º“wound up” 6-story Scherk towerWhat would happen,if we added more stories ?or introduced a twist before closing the ring ?
40“Hyperbolic Hexagon II” (wood) ISAMA 2004“Hyperbolic Hexagon II” (wood)For whom I designed certain shapes on the computer which he then built oin wood.Brent Collins
41The Generative Process Find the inherent constructive logic.Devise an appropriate generative program.Introduce sliders for crucial parameters.Play with sliders to explore design space.Reprogram to go outside current domain.Think outside the box !Many, many experiments . . . The computer becomes an amplifier for an artist’s creativity !
42Snowsculpting Championships 2003 Silver Medal Winner: “Whirled White Web” (C. Séquin, S. Wagon, D. Schwalbe, B. Collins, S. Reinmuth)