Patch loading : resistance of longitudinally stiffened steel girder webs

Sammanfattning: In most cases, longitudinal stiffeners increase the resistance of steel girder webs subjected to concentrated forces, commonly referred to as patch loading. During the last three decades, a number of experimental investigations have been undertaken in all over the world in this regard. The test results shows that the patch loading resistance increases when a longitudinal stiffener is placed rather close to the loaded flange, however very little has been done to quantify this increase which in most design codes has been disregarded. At first the effect of a longitudinal stiffener on the structural behaviour of plate girder webs subjected to patch loading is investigated by means of nonlinear finite element analysis. Two plate girders with extreme values for the flange-to-web thickness ratio were modelled taken into account material and geometrical nonlinearities. The numerical results were validated with test results from others. Thereafter, a parametric study was conducted in order to investigate the relevance of some parameters such as the relative position and flexural rigidity of the stiffener and the size of the loaded flange in increasing the ultimate resistance of the steel girder webs to patch loading. In the past, it has been customary to calculate the patch loading resistance of the girder web with a longitudinal stiffener by multiplying the corresponding resistance of the unstiffened girder web (web only with transversal stiffeners) with a correction factor, obtained by a regression analysis, depending mainly on the relative location of the stiffener. Herein, a sensitivity analysis was conducted showing also the relevance of the flange-to-web thickness ratio and the ratio of the yield strength of the flange to the yield resistance of the web. Accordingly, a new correction factor was proposed taken into account the aforementioned parameters. The failure mechanism of longitudinally stiffened girder webs subjected to patch loading observed in the experiments is similar to that of unstiffened girder webs. This is characterised by the presence of elastoplastic deformation in the patch loaded flange (plastic hinges) and in the web (yield lines). From this deformed configuration, a mechanism resembling the distortions in the web plate and loaded flange was developed in order to obtain the patch loading resistance of longitudinally stiffened girder webs. This failure mechanism model fails in that the actual material and geometrical properties of the stiffener are not taken into account and certain assumptions are accordingly made to overcome this situation. In the latest edition of the Eurocode 3 Part 1.5 the patch loading resistance of unstiffened girder webs is obtained using a resistance reduction factor applied to the yield resistance to patch loading. This factor depends on a slenderness parameter, which is the square root of the ratio of the yield resistance to the critical buckling load (according to classical theory of elasticity) for patch loading. However, the interaction with longitudinal stiffening is not considered. In some respect the use of a longitudinal stiffener is comparable with a reduction in the slenderness ratio of a girder web. An expression is found for the buckling coefficient used for calculating the critical buckling load for patch loading based on finite element analysis. The interaction between the web plate with flanges and a longitudinal stiffener (open and closed section) was considered in the numerical analysis. Consequently, the critical buckling load increases due the presence of the stiffener and then the slenderness parameter is reduced. This leads to an increase in the patch loading resistance. Finally, the results obtained using the methodologies described above and those found in the literature are compared with available tests results. The goodness and flaws of each methodology are further discussed.