Elastoplastic large deflection analyses of simply supported thick rectangular plates with initial deflection have been performed using nonlinear finite element method. The objective of the study is to check to what extent the plastic correction method used in the present classification rules is reliable for estimating buckling strength of a thick rectangular plate under combined in-plane loads. The buckling strength interaction curves of classification rules are compared with the ultimate strength interaction curves obtained from FEM analyses, and the degree of margin of safety is examined. A simple modification of plastic correction method is proposed for some combined loading cases for which it is revealed that the present plastic correction method estimates the buckling strength on the unsafe side.
Hull girder strength of a ship is derived from a lot of rectangular plates strengthened by stiffeners. These plates axe subjected to many in-plane loads such as compression, shear, bending etc. and in the design of such plates it is inevitabh to check their buckling strength. All the classification societies have prescribed standards for the evaluation of buckling strength of plates. At present, buckling strength evaluation of classification rule consists of estimating elastic buckling strength from elastic buckling interaction curves and then applying plastic correction using Johnson's formula etc. for the plates where plastic buckling may occur. A lot of researches have been done on elastic buckling interaction and reasonably good results have been obtained in this regard. On the other hand, plastic correction formulae axe originally proposed as the safe side approximation of ultimate strength for column members. Although these formulae have been applied for plates straightforwardly, it does not have any physical meaning in itself. Also the strength of actual "plates depend strongly on initial imperfections.
