Phase 003
-framework with edge panels assembled
-release application applied onto mold
-frame positioned and fiber composite is poured (2 cups per panel – 5mm fiber)
Phase 002
-panels are laser cut from 6.5mm ply (2 frames per half sheet)
-assembled panel is placed into jig, secured, and check for angle deviations
-frames are mounted into place verifying global curvature
Phase 001
-mold bases are laser cut and assembled
-revolver template is water jet and mounted to revolution head. Mold and head are then calibrated
-plaster is then applied in layers and smooth until class 9,10 surface is met… repeat
Development of the Virtual Anechoic Chamber (VAC) being explored in C#/GH/Processing/Flex-in/Flex-Out
This project requires the creation of digital parametric models that define the geometry of the hyperbolic surfaces.
Different options of these surfaces can be tested digitally, using computational acoustic analysis techniques suitable for the prediction of sound scattering, and tested physically, using scale models and testing equipment.
The primary research question is: What are the acoustic properties of doubly-ruled surfaces?
1. The acoustic performance of a surface can be modified through geometry or material.
2. When sound strikes a surfaces it is absorbed, reflected, or scattered.
3. This research project is interested in using a single material – plaster.
4. This aim of this project is to determine the acoustic properties of a particular geometry – doubly-ruled surfaces.
5. As we are only interested in a single material, and it is a hard, reflecting material, then we are not interested in absorption.
6. As the geometry we are interested in testing is complex and comprised of relatively small faces, then we are not testing the reflection.
7. Then, the acoustic performance attribute that we are interested in studying is the sound scattering / sound diffusive properties of doubly-ruled surfaces.
We would like to welcome the following participants selected for the Responsive Acoustical Surfacing Cluster at SG2011:
Participants:
Adam Laskowitz
Amir Gazit
Ben Coorey
Dusanka Popovska
Giovanni Betti
Kathy Yuen
Ralf Lindemann
Robin Bentley
Ruwan Fernando
Thomas Hay
Champions: Phil Ayres – Architect, Educator, Researcher at CITA, Royal Danish School of Fine Arts
Mark Burry – Professor and Director of SIAL and Design Research Institute (DRI), RMIT University
Jane Burry – Research Fellow, SIAL, RMIT University
Daniel Davis – PhD candidate SIAL, RMIT University
John Klein – Architect, Computational Designer, Zaha Hadid Architects / SCI-Arc
Alex Pena de Leon – PhD candidate SIAL, RMIT University
Brady Peters – PhD Fellow at the CITA, researcher with JJW Arkitekter / Grontmij / CarlBro Engineers
Support: Peter Holmes – SIAL Sound Lab, Acoustician based in Melbourne
Brad Marmion – SIAL modeling lab coordinator
Individual cluster contact info and resources are under the participants and reading pages.
In Galicia (2000), our stonemasons worked by making 1:1 mock-ups of the individual stone pieces by building up layers of 10cm polystyrene.
I’ve attached a couple of drawings, from a set of over 700 that Jane and I produced for the purpose.
The blue lines are 10cm contours for the polystyrene cuts.
Stock in place.
Fixed dimension between driven portals is 1420mm. Operating limits for portals is 500mm for y-axis (depth), and 300mm x-axis (height).
Straight fluted column – easy to remove from stock. Far right: Hyperboloid – less easy to remove from stock.
Fluted hyperboloid – complete pain to remove from stock. Right: stock after removal – a better finish than the positive.
The interface (MS-DOS). Simulation of the procedure. Portal drive geometry projected from the cut geometry on the stock face.
NOTES:
1. Cutting speed is relatively quick – about 3mins for the fluted hyperboloid.
2. The proportions between the driven area of portals and the fixed 1420mm dimension between portals favours the long and slender over the short and squat.
2a. It would be feasible to add a revolving base to resolve this . Based on the testing above I think this could easily be rotated and controlled manually.
Takes about 30 minutes to mix plaster and pour + 1 hour to set, although it is probably a day before it is full strength.
Mix the plaster the to instructions on the bag. If you need it thicker, don’t add more plaster, just wait until the plaster starts to cure.
Pour the plaster slightly before you think it is ready, because you can always push the plaster back to the top, and the curing process happens really fast.
Pouring too much plaster gunks up the revolver, add it bit by bit.
Make sure the revolver has mold release on it.
Vaseline is a suitable mold release.
Forcing compressed air into the mold release separates the cast from the mold.
1. There are two main pours: the first bulks the framework up, the second is the finishing surface.
2. Make the bulk from firm, dry plaster and use a smooth, runny plaster for the finishing surface
3. Let the plaster dry prior to the final pour or else gravity will pull everything down.
4. Any lumps in the final mix catch on the edge of the revolver and scar the surface.
5. Keep the edge of the revolver clean – make it so the edge of the revolver is easy to clean.
6. Takes about an hour to make one hyperboloid.
GlueLess Revolving Pyramid Station
Reduces the Interior Volume of the Hyberboloid Molds, without this we would have to use the total volume of the shape.
in the center of the base a rotation axis tube will be located… from it we will revolve the laser cut profiles.
