By dotA
Hornoarable Mention
Team: Chang Qiang, Gao Yan, Duo Ning, Peter Chan Shu Kei
BOW is a pavilion computed by repeating bespoke components, of which each unit comprises a straight acrylic extrusion and a bended aluminum extrusion (此处见左上角的图), working together as a pre-stressed structural unit to improve the stiffness and overall strength of laminal material. By changing the width of the bended aluminum sheets to fit with the surface curvature, a continuous folded 3d twisting surface can be achieved (此处见repeat-cq-finalconstruction-20100930.dwg中的三个从平面到任意曲面的演变图).
A very simple notch joint (DIA_01) is used for connection, which is totally re-demountable and fairly cheap. The joint can be fabricated while cutting the planar strip, with only laser-cut or 2/3D CNC machine. The shape of each component is determined by the positions of the notch joints and the shape of the plane-cut aluminum sheet. At joints, the acrylic strip and the aluminum sheet are stiffened each other to form a connection with high stability of mechanics of a thin-wall structure. The notch joint can also be constructed easily without involving complicated manufacture and construction facilities.
The aluminum sheet material can be as thin as 1~2mm when it is bended. And the straight acrylic sheets are also stress-stiffened by the pre-stress.
The overall geometry is the result of 3 intersecting elliptical curves by cutting 3 cylinders inspired by the min/max circular constrains of the site with rotating planes, which are set up according to orientation, access, visuality, as well as component assembling logic (DIA_03).
The whole structure concept is based on 3 grid-type twisty-surface beams defined by the 3 overall intersecting elliptical curves. These beams support each other to ensure the global stability. The height of the beam reduces the stress of overall bending. While pre-stress in the components can resist these stresses to keep the whole structure tight.
BOW is a complex 3D assembly made out of simple 2D components by the design intelligence rather than the machine intelligence.
