The optimisation problem related to the tessellation of free-form grid shells is presented in this paper. This kind of structure is generally composed of a supporting grid that defines the geometry of a large number of cladding elements that are always different from each other. From the construction point of view this means that each single piece needs to be designed and produced “ad hoc”, then marked and positioned with the aid of an assembling table. In order to reduce the heterogeneity of grid shell elements, several optimisation strategies, referring both to evolutionary and gradient-based techniques, have been tested and compared. All the free-form geometries are defined and handled with a commercial NURBS based software, while the development of all the optimisation procedures has been made thanks to the VB based programming language implemented inside the NURBS based software. Due to the smoothness of the solution domain of this specific problem, gradient based procedures seem to be the most efficient for the rapid convergence to the optimal solution. Finally a multi-objective procedure, that involves static analysis combined with the discussed geometrical optimisation, is proposed in view of future developments.
Computational Morphogenesis in Architecture: the cost optimization of free-form Grid Shells
DEL GROSSO, ANDREA;
2009-01-01
Abstract
The optimisation problem related to the tessellation of free-form grid shells is presented in this paper. This kind of structure is generally composed of a supporting grid that defines the geometry of a large number of cladding elements that are always different from each other. From the construction point of view this means that each single piece needs to be designed and produced “ad hoc”, then marked and positioned with the aid of an assembling table. In order to reduce the heterogeneity of grid shell elements, several optimisation strategies, referring both to evolutionary and gradient-based techniques, have been tested and compared. All the free-form geometries are defined and handled with a commercial NURBS based software, while the development of all the optimisation procedures has been made thanks to the VB based programming language implemented inside the NURBS based software. Due to the smoothness of the solution domain of this specific problem, gradient based procedures seem to be the most efficient for the rapid convergence to the optimal solution. Finally a multi-objective procedure, that involves static analysis combined with the discussed geometrical optimisation, is proposed in view of future developments.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
