ABSTRACT

ABSTRACT: In the history of construction, shell structures always represented a conceptual challenge for architects and engineers: from ancient masonry vaults to reinforced concrete roofs, their intrinsic three dimensional behavior, in terms of stresses and failure mechanisms, can be hardly predicted by intuition, because it relies on the complex interactions between geometrical and mechanical properties. In the past the conception was ruled mainly by geometrical proportions or trying to apply the knowledge on mechanics of arches; the diffusion of reinforced concrete and the introduction of experimentation on physical models, in the first half of the twentieth century, brought to the golden age of Tedesko, Torroja, Candela and Isler. The complexity of partial derivative problems related to shells created a gap between theory and practice until the sixties, and the development of computational technology and numerical methods did not revolution immediately this field as it could be expected, focusing in the beginning rather on analysis and behavior prediction than on conception. Only recently the development of optimization techniques, numerical form finding and generative approaches supported by computing seem open new scenarios to designers.