元結, 正次郎MOTOYUI, SHOJIRO
, p.791 , 2016-04 , 日本建築学会
Composite beam in a frame subjected to story drift consists of two portions of positive and negative bending moments, respectively. The lengths and moment magnitudes of the two portions are different due to the composite action and increased bending stiffness caused by positive bending. Thus, in order to understand the composite beam behavior, it is necessary to consider interaction between the two portions. This paper, therefore, proposes a practical analysis method to simulate in detail the behavior of a steel beam, a concrete slab, and stud connectors, and uses it to clarify the interaction as well as the composite action of the beam in double curvature bending. Two types of composite beam models are used. One consists of multi-spring element, truss element, and beam element. Another is a FEM model that simulates nonlinear behavior of both steel and concrete. Both models simulate slip between the concrete slab and steel beam, bending of stud connectors, compression and tension failure of concrete slab, contact and separation between the slab and beam end, and yielding of the steel beam. Analysis results are compared with the past test result of composite beam subjected to double curvature. Two analysis models reproduced initial stiffness, strength, and moment-rotation curves of the test result with excellent accuracy (Fig. 11 and Table 7). The analyses also gave more insight into the complex behavior of the composite beam subjected to the double curvature bending of various magnitudes, which are discussed below: The steel beam of beam ends subjected to the same rotation, the beam end under positive bending and that under negative bending yielded almost at the same time (Fig. 15). Shear force of a stud connector in a beam portion of positive bending grow larger when the steel beam yields there. On the other hand, it decreased when the concrete slab crashes there. The yield zone of the steel beam was concentrated into the bottom flange the case of positive bending, and it was over all the cross section in the case of negative bending (Fig 18). From elastic loading to inelastic loading exhibiting the beam rotation θ=0.02, bottom flange strains at both ends of the beam were about equal (Fig 19). In case of composite beam subjected to double curvature, a ratio of bottom flange strains of the two beam ends under positive and negative bending, respectively, appears to be proportional to inverse of the ratio of moment inertia, the ratio of neutral axis distance from the bottom flange, and a ratio of portion length (Eqs. 5b). From elastic loading to inelastic loading exhibiting the beam rotation θ=0.02, the bottom flange strain ratio is about 1.0, which agrees with the analysis result. This cause is because the beam deflects relatively less under positive bending, and beam curvature grows in negative moment. The effect of rotational restraint provided by the columns at the both ends of the beam is examined. Because the restraint is larger at the exterior column where only one beam is connected, the bottom flange strain tends larger. However, at θ=0.02 and larger, curvature increase of the end under negative bending is remarkably large, which limits strain of the positive bending side. The exterior column case mentioned above as well as a single span case show that bottom flange strain of the beam in positive bending and thus composite action tends to be smaller than in negative bending.