抄録
The importance of precaution against unexpected earthquakes was highlighted after the 2011 Great East Japan Earthquake. While Nankai trough mega quake and a Tokyo inland earthquake are expected in the near future, it is important to prepare for disasters in metropolises and other urban areas. Since high-rise steel buildings in urban areas are the basis of civic and business life, adequate methods of evaluating their structural soundness, which enable an immediate judgment of the prospect of business continuity and the early recovery of such buildings after the disaster, are required. For this purpose, it is necessary to clarify the collapse behaviors of such buildings.
A number of experimental and analytical studies on collapse behavior of steel moment resisting frame have been conducted by many researchers. However, most specimens were a few stories models, which are much smaller than actual buildings. There have been no shaking table tests simulating the entire high-rise steel moment frame. It is necessary to clarify the relationship between the extension of damages of the components and the collapse behavior by large scale shaking table test with many-story frame specimen.
This paper reports on a large-scale shaking table test of a high-rise steel building conducted at the facility called E-Defense. The building specimen is a 1/3-scale 18-story steel moment frame designed and constructed according to design specifications and practices used in the 1980s and 1990s. The input motion is a long-period and long-duration strong ground motion simulated considering a Tokai, Tonankai and Nankai coinstantaneous earthquake. The specimen was subjected to a series of progressively increasing scaled motions until it completely collapsed. At first, the yielding was observed at the beam-ends along the lower stories and the column bases of the first story. As the ground motions became stronger, cracks initiated at the welded moment connections, and then fractures spread in the beam flanges of the lower stories. As the shear strength of lower stories decreased owing to such damage, the story drifts significantly increased and the frame finally collapsed. The main cause of the collapse of the test specimen was the extension of the fractures of the lower flanges at the beam-ends of the lower stories. Furthermore, the changes of the natural period, damping ratio, and participation vectors were calculated from the test result data and the relationship between the changes and the extension of the structural damages like local buckling or fracture was presented.
In addition, a numerical analysis was performed in advance of the shaking table test. The deterioration hysteresis model was used which can represent the characteristics after fracture of lower flange at beam ends. The analysis results were roughly similar to those of the test in terms that the analysis model exhibited the fractures of the lower flanges at the beam-ends of the lower stories and then the significant deformation, as observed in the shaking table test.
