Comprobación experimental de las cargas de pandeo de planchas con refuerzos

Abstract

Due to the limited experience in the ecuadorean scope on estimation of critical loads of structures formed by plates with reinforcements, this project sought to revise formulations to evaluate the mentioned parameter, and compare it with experimental results. The classic reference [7], describes four typical buckling modes for such structures: plate between 2 reinforcements, Flexural bending of stiffeners, torsion of stiffeners, and Total deflection of the panel. Formulations were implemented for each buckling mode. The reviewed formulae for estimating critical load for buckling of reinforced panels were implemented, and after varying the geometric parameters were set four panels, each of whom had at least the effort of one of the critical modes of buckling and described. Then, using stainless steel plate 1.5 mm thick, were built two specimens of each type of panel, with a maximum length of 1.20 m, considering the available equipment. Before developing the test, thickness and the resulting pre-deformation were measured. Thickness had little variation but pre-deformations were high, in the order of the thickness of the plate The experimental tests were conducted in the Universal Testing Machine of FIMCP, and determined the maximum force that bore the specimen. On the buckling of the plate between stiffeners, the experimental results are larger than the analytic, because the edges of the plate are welded to the reinforcements, while the analytical model considered simply supported. On the Flexural Buckling of longitudinal reinforcement, the theoretical results are higher than experimental ones, because of local failure between flange and web of the reinforcements, the distance between welds. In the torsional buckling, the plate twisted instead of the stiffeners, because side edges were not stiffen enough, and this substantially reduced the experimental effort critical. In total buckling of the panel, the longitudinal reinforcements flexes in parallel, since the edge stiffeners were not enough, resulting in a substantial difference between the results.