Kuşatılmış yığma yapıların deprem davranışı: 2011 Van depremi örneği

Masonry structures exist in most earthquake prone countries, despite the non-ductile behavior of most unreinforced masonry (URM) types. Several earthquakes around the world proved, however, that if necessary precautions are taken, masonry structure can also behave well during strong seismic events. One of the ways of improving the seismic resistance of the masonry buildings is to construct confined masonry structures including properly detailed vertical and horizontal lintels. Confined masonry does not exist in the Turkish seismic codes as a building type, as of 2016, thus the design issues are not addressed. Nevertheless, there are several confined masonry structures built throughout the country and, as observed during the 2011 Van earthquakes, they did behave well in overall. In this paper, two real structures taken from the city center of Van, eastern Turkey, have been examined. These structures were built by using the local techniques, i.e. constructing a stone foundation approximately one-meter-high, embedding vertical RC elements into that, and then building the rest of the structure as an amalgamation of reinforced concrete and masonry. The main difference of the construction from the widely seen RC buildings with masonry infills comes from the fact that in the confined masonry the walls are constructed before the RC frame, thus the RC members are fully attached to the masonry. This certainly alters the loading paths, and allows masonry walls participating in the vertical load bearing capacity, leading thus to higher shear load capacities of the walls as well as to increased contact between the masonry walls and the RC frame. Nonlinear time history analyses have been conducted on the two case-study structures found on the site visit after the 2011 Tabanlı, Van Earthquake. Geometrical dimensions, reinforcement amounts as well as the damages have been recorded on site and the structures have been modelled by using the site data as well as assumed data regarding the masonry strength. Analyses have been conducted once on a bare frame where RC members have the same dimensions and reinforcement with that of the confined masonry structures. Another set of analyses has been conducted assuming that the very same masonry walls were constructed as infill walls, meaning that they are not fully attached to the structure and do not take any portion of the total vertical load. Finally, the structures have been modelled as they are by using the confined masonry properties. The same material properties have been used in infilled and confined masonry models so that the comparison is decoupled from the masonry material properties. In the modelling, distributed plasticity force-based elements are used, while the masonry walls have been modelled by using the double-strut modelling approach implemented into the software used. Unfortunately, there are no acceleration records available from the site, because of malfunction of the recorder in the Van city center. In order to simulate what might had happened in the center, the closest record of Muradiye Station has been processed and it was found that the median prediction of the Abrahamson and Silva (2008) model matches quite fairly with the fault-parallel and fault-normal components of the Muradiye station. Taking advantage of that, the Muradiye record has been scaled up to match this time the predicted spectrum in the Van city center. These new scaled records have been used in the analyses. The results have been examined in terms of maximum base shear attained during the analyses, and the total area under the hysteresis loops. It was found that the confined masonry structures have 33% higher strength and 46% higher hysteretic energy consumption in average as compared to their infilled counterparts. Additionally, the confined masonry structures exhibited approximately 3 times higher base shear capacities than the bare structures. Finally, the damage state observed in the Case Study 2 was found in agreement with the site observations, meaning a reparable level of moderate damage.