Popis: |
Stiffened plates are the most common structural components used in many engineering branches such as aircraft, trains, ships, offshore structures, and buildings of every kind and dimension. In the marine field, in particular, merchant and naval ship hulls can be considered to be an assembling of simple modules such as bottoms, sides, decks and bulkheads, all formed by stiffened plates. This very simple structural model has the advantage of offering a good compromise between strength, construction cost and weight. At the same time, the scantling of a stiffened plate can easily be obtained by hand calculations or by more refined tools, such as FEM codes, if a more reliable optimization is required. If the first approach is used, the first action to be performed is the calculation of the plate width acting with the stiffeners, usually referred to as “effective breadth”. This aspect, widely assessed in the steel structural field since the 50s, still needs further investigations in the field of composite materials, commonly used for building motor and sailing yachts. Given the present financially critical situation affecting yachts, especially in the range of small and medium size vessels, many shipyards, with a long experience in the GRP construction, devote more attention, on one hand, to significantly reduce the weight and cost of the hull structure and, on the other hand, to increase the vessel dimensions to improve their market potential. In both cases, a more detailed structure scantling is required and, in the case of a “by hand” calculation, this mainly implies a deeper knowledge of the “effective breadth” concept for composites. In this paper, a numerical procedure for the calculation of the effective breadth of GRP, single skin and sandwich stiffened plates is presented, using a numerical approach based on the finite element method. This study is focused on plates stiffened by GRP box beams for single skin panels and by GRP girder beams for sandwich ones. All the geometric characteristics of the model have been varied parametrically, in order to obtain a very wide data series that has been statistically processed. |