In the 2011 off the Pacific coast of Tohoku Earthquake, many suspended ceilings and other suspended equipment fell down due to the lack of their resistance to earthquakes. To mitigate severe damage to ceiling system caused by earthquakes, new seismic design code for suspended ceiling system was issued by the Ministry of Land, Infrastructure, Transport and Tourism. However, the mechanism why and how suspended ceiling system falls down during earthquakes has not yet been clarified well. In order to clarify the collapse mechanism of wide-area ceiling system and development of its countermeasure, new research project was launched and the first series of full-scale shake table experiments of wide-area ceiling system in school gymnasium was conducted. This paper presents outline of the full-scale shake table experiment and global response of structural members.
The specimen was designed as the full-scale specimen which can represents real steel school gymnasium on spread foundations built in elementary and junior high schools based on the allowable stress design with base shear coefficient C0 of 0.2. It had a floor plan dimension of 30m by 18.6m and a height of 9.09m. Because its size was larger than the shake table size of 20m by 15m, it was supported by cantilevered large stiff girders with the overhang of 5m at the maximum. Based on pushover analysis using Ai distributions, it was confirmed that the base-shear versus displacement relation in the specimen was close to that in the prototype of the specimen. In the specimen, two different types of suspended ceiling were installed; 1) non-seismic ceiling and 2) seismically designed ceilings with seismic coefficient of 1.1G and 2.2G. Fail-safe system consisting of wires and nets to prevent damage caused by suspended equipment to people inside the gymnasium was also installed to evaluate its effectiveness.
Two ground acceleration records were used as imposed motions; 1) K-NET Sendai record observed at K-NET Sendai station during the 2011 off the Pacific coast of Tohoku Earthquake and 2) JMA Kobe record observed at JMA Kobe observatory during 1995 Hyogo-ken Nanbu earthquake. Intensity of imposed motions were 5%, 25% and 50% (twice) of K-NET Sendai record for the specimen with non-seismic ceiling and 5%, 25%, 50%, 80% and 100% of K-NET Sendai record and 100% and 150% of JMA Kobe record for the specimen with seismically designed ceilings.
Based on the shake table experiment, it was found that response accelerations measured at the base of corner columns (the edge of cantilevered stiff girders) were close to the original record in horizontal directions while vertical response accelerations were amplified in the frequency of 10Hz. It showed the response acceleration should be carefully checked in high frequency components.
Based on the whitenoise excitations, natural periods and mode shapes were estimated. It was clarified that the vertical response was coupled with horizontal response and the intensity of vertical response was the maximum value of 32% of the intensity of horizontal response.
Based on the excitations using K-NET Sendai and JMA Kobe records, the maximum response displacement at the roof top was 2.05% (=1/49) and 4.47% (=1/22) of the height of specimen in span and ridge directions, respectively. Only the base of columns yielded, however, no damage was observed in columns, girders and beams. All vertical and horizontal braces yielded and buckled repeatedly, and the deformation in the out-of-plane direction was observed. However, no brace ruptured.
