DAMAGE OBSERVATION AND DETECTION USING ACCELEROMETER OF DRY PARTITION WALL INSTALLED IN STEEL MOMENT FRAME UNDER LARGE STORY DRIFT

From a viewpoint of continuous use of buildings after an earthquake, it is important to understand the behavior and damage progress of non-structural elements such as dry partition wall under a large deformation owing to earthquake, and to establish a method to monitor the condition of non-structural element in a qualitative manner as well as in a quantitative manner. In this paper, first the past studies on loading experiments for dry partition wall were summarized, showing that enough studies considering the effect of opening and the behavior of walls located away from beam location have not been done. Based on the review of past studies, a full-scale static loading test of steel moment frame with dry partition walls was performed. The specimen has two frames; one frame has a flat wall without opening and the other frame has a wall with a door opening and a corner in the frame. In the latter frame, some part of the wall is placed not directly under the beam. Static cyclic loading was applied, increasing the peak story drift up to as large as 1/33. In the experiment, visual observation was performed at each loading step. Rotation angle and vibration of the walls was also measured using MEMS acceleration sensors. The flat wall had relatively few board cracks and other damage. However, the boards on upper layer in the center part of the wall began to rock at the loading step of a 1/100 story drift. Then at 1/50, the boards showed out-of-plane deformation and when unloading after reaching 1/33, boards fell off. On the other hand, on the wall with the opening, a gap opening between the boards, and a crack on the L-shaped boards at the corner of the opening were observed at a relatively small story drift angle of 1/400. Regarding the wall located away from the beam, damage was concentrated on the wall which was orthogonal to the frame and not under the beam. The wall rotated around the vertical axis, and this was thought to be due to the difference of trackability to the structure between walls. Due to this behavior, a crack at the corner of the wall was seen at the story drift angle of 1/400, and at 1/75 the upper track opened, and studs were seen to come off. The rotation angle and the vibration of the walls were measured by MEMS acceleration sensors. The measurement result corresponded well with the damage observation. It was found from the measurement results that there were signs before damage and behavior of walls could be visually confirmed. Moreov