Stressed Skins explores how very thin, easily bent metal sheet can become a strong but lightweight structure. Architects use thin metal sheets as cladding panels to provide integrated enclosure, structure and form. Because loads vary over such a building system, performance requirements vary, and customized load-adapted panel designs could mean significant efficiencies of material use and possible reductions for supporting structural systems.
This project develops workflows and methods to support customised design and fabrication using Incremental Sheet Forming (ISF). These include the prediction of changes in material properties such as thinning and work hardening, the automated generation of load adapted rigidisation geometries, the prediction of overall structural behavior, and the automated generation of fabrication information. A specific concern is the development of adaptive mesh-based methods as a means to communicate information about design, material properties and performance across scales.
Team: Paul Nicholas, David Stasiuk, Esben Clausen Nørgaard
Materials Testing: Monash University, Materials Science and Engineering: Prof Christopher Hutchinson
Structural Engineering: Bollinger + Grohmann: Robert Vierlinger, Clemens Preisinger
Robotics: Thibault Schwartz, HAL Robotics ltd
Photography: Anders Ingvartsen
STRESSED SKIN STRUCTURES
Stressed Skins is a frameless, stressed-skin structure, with tension, compression and shear forces carried through the skin. In the design of stressed skins, one of the main problems is to ensure rigidity at multiple scales: against the instability of the whole structure and also the local buckling of the parts which have to carry compressive load. Stressed Skins develops a structural approach in which local corrugation resists buckling through geometric stiffening of the skin, while shear connectors transfer loads between upper and lower skins to rigidise the entire structure.
ROBOTIC INCREMENTAL SHEET FORMING
Stressed Skins explores the transfer of forming technology from automotive into architecture. Each panel is formed by a robot directly linked to a 3D computer model, a process known as Robotic Incremental Sheet Forming (ISF). In this fabrication method, a ball-head tool is moved over the surface of a thin flat metal sheet, progressively imparting localised plastic deformation and 3D form. This stretching of the metal has implications for material properties: the sheet ungoes localised thinning and hardening, and can no longer be considered as having uniform properties. A key contribution of Stressed Skins is to develop a design-integrated modeling process that connects calculation of these forming implications to performance at material, element and structural scales.
The modelling approach used for Stressed Skins is necessarily multi-scale. At the macro scale, it is important that the design is aware of possible instability in the whole structure, at the meso scale, of local buckling of the panel elements, and at the micro scale, of material thickness and the potential for micro buckling during the forming process. These parameters are to a large extent interdependant.
The multi-scale modelling approach used in this research project is comprised of techniques which enable the information generated at each scale to flow both up and down the continuum. Here, an adaptive mesh refinement method is used to support localised variations in resolution and information flow. From the perspective of the design development, these include: overall form-finding and panelisation operations; global structural analysis and adaptive specification of connectivity arrays; and recursive local tectonic pattern formation which depends upon finite element analyses and is further informed through the calculation of forming strains and material thinning.
FABRICATION AND TOOLING
Panels of approximately 50*100 cm were fabricated using an ABB industrial robot. Rather than using conventional toolpath generation algorithms, in which toolpaths are generated as sliced contours in the horizontal plane, a custom spiral-based algorithm was developed that included grouping of features, recofiguration of jig positioning to fit optimimal zones of forming, and adjustments to tooling speed with relation to wall angle.