Local Buckling Properties Of Stiffened Plates In Steel Box Girders

In this paper, local buckling properties of stiffened plates in steel box girders were studied. The representative mechanical models were submitted to uniaxial compression and simply supported boundary condition.To begin with, linear-elastic buckling theory was analyzed since it had been found considerable application in the solution of engineering problems. The orthotropic plate approach and the energy principle were applied to calculate the buckling stress, respectively. Then a comparative research was conducted. It was concluded that the orthotropic plate approach was far from accuracy because of the rough equivalent stiffness while the energy principle was a wise choice to evaluate the buckling stress based on the linear-elastic buckling theory.Furthermore, the present study developed a simple analytical method to caculate the ultimate strength of stiffened panels. Based on the nonlinear theory, the presented method proposed to take the effects of residual stresses, imperfections, material and geometric nonlinearity into consideration when aiming to analyze the ultimate strength of steel panels stiffened by open or closed ribs. Because it was found that the residual stress and geometric nonlinearity seemed sensitive to the stability of steel plate panels while the magnitude of initial geometric imperfections did not. However, the pattern of initial geometric imperfections significantly affected its failure modes. Two specimens from the compression tests were presented and their ultimate strengthes were proved to be well predicted by the proposed method in this study afterwards.At the same time, the ultimate strengths of the two specimens mentioned above were also evaluated by three bridge specifications:the British Standards 5400(Abbr. BS5400), the American Association of State Highway and Transporttion Officials' Load and Resistance Factor Design Specification (Abbr. AASHTO LRFD) and the Japan Road Association's Specification for Highway Bridges. Findings from code comparisons showed that:(1) the BS5400 seemed significantly conservative; (2) the measured ultimate strength was better predicted by AASHTO LRFD; (3) the JRA specification slightly overestimated the ultimate strength of both specimens.In addition, this paper presented an optimum design of longitudinal stiffeners. An example was given to illustrate the applicability of the equation and comparative analyses were also made to demonstrate the reliability of the analytical study conducted.