繰り返し変位履歴を受ける梁端溶接部の塑性変形能力評価

Abstract

In Tohoku earthquake (March 11, 2011), highrise steel buildings of the Tokyo downtown area shook for a long moment under the influence of long-period ground motion. To prevent damage under such ground motion, evaluation on deformation capacity of steel member (welded beam end, etc.) has been required under multi-cycle loading. In this paper, deformation capacity of the welded beam-end is assessed by the FE analysis, considering fracture rule, which adopts both cyclic damage rule and monotonic damage rule. The cyclic damage rule is a fatigue damage law based on Continuum Damage Mechanics (CDM), and the monotonic damage rule is a damage rule for a large ductility factor equivalent to monotonic loading. The validity of this fracture rule is verified by the simulation of past experiments under cyclic loading. Two past experiments are selected. One is an element experiment using specimens modeling beam flange and web, in which tensile and compressive repeated force is applied in the axial direction. The other is a partial frame experiment using field welding type specimen having a ¼ circular weld access hole of R23 at the beam-end, in which increasing cyclic force is applied. For an element experiment, simulation analysis of monotonic force and constant amplitude repeated force is performed using FEM with solid elements and shell elements. As a result of the FE analysis, we can simulate the reduction of peak load, and it is found that the deformation capacity under multiple cyclic loading conditions can be evaluated. For a partial frame experiment, simulation analysis using shell elements is carried out. As a result of the FE analysis, it is found that a crack occurs at the toe of the weld access hole on the lower flange, and propagates in the flange width direction. Using FEM it is possible to simulate the situation where the welded beam-end fractures, and it has been verified that the load and deformation relationship of FE analysis had good agreement with test results. In addition, we carries out a comparison between numerical results and the performance curves of the welded beam-end proposed by MLIT project (the project name; 27-2) and JSSC committee. It is demonstrated that a performance curve is possible to be evaluated considering effect of beam-end details using the proposed numerical method.