Statistical evaluation of the fragility of existing RC buildings in turkey under seismic loads

Turkey had been exposed to several devastating shakes throughout the last quarter century, which had exhibited the insufficient capacity of the existing building stock under destructive earthquakes. A clear majority of the total building stock in Turkey consists of low- and mid-rise reinforced concrete (RC) frame structures. The general run of these structures are non-engineered or have not been adequately designed and/or constructed according to the then-current earthquake code regulations. Various deficiencies that the existing building stock have, caused severe damages and partial or even total collapse which have eventuated in severe socioeconomic losses during the recent earthquakes of Turkey. These defects mainly depend on poor structural material quality, inadequate reinforcement detailing, lack of confinement zones, heavy and large-span cantilevers and indirect supporting preventing the formation of regular structural frames due to the fact that there is a lack of control mechanism during the construction stages for most of the structures constructed prior to the National Construction Control and Supervising Law. Therefore, designating the seismic reliability of existing building stock specifically located in the moderate and high seismicity regions and performing damage/loss assessments for these locations are important necessitations in the mitigation of losses during future seismic events. Fragility curves offer the probability of reaching or exceeding a level of damage for an individual building as well as building stocks under a ground motion intensity parameter by means of earthquake intensity-damage relations. They are very handy tools which can be utilized for pre-earthquake planning; strengthening intervention; damage/loss estimation and disaster management plans to be drawn up. Furthermore, when considering the difficulty on making a decision about the whole building stock, they serve as easy and fast applications when they are applied to estimate the probable seismic damage for building stocks settled on a seismic zone. Although a fragility curve offers the probability of exceeding a predefined structural damage limit in terms of a ground motion intensity parameter: i.e, peak ground acceleration or velocity (PGA or PGV), elastic spectral acceleration or displacement (Sa or Sd), it is observed that the level of damage state can be related to the structural capacity and demand. Other than the structural system properties, however, the post-earthquake damage state of a building is dependent to many parameters including local site conditions; structural material qualities; adequacy of the workmanship and detailing of reinforcement. The need to predict the vulnerability of existing structures has led to increase the researches concerning with the development of seismic fragility assessment methods. Existing fragility functions can be classified into four generic groups of empirical, judgmental, analytical and hybrid functions according to whether the damage data used in their generation derives mainly from observed post-earthquake surveys, expert opinion, analytical simulations or combinations of these (Rosetto and Elnashai 2003). Analytical fragility relationships can be established either by employing a set of nonlinear dynamic or pushover analyses. In this dissertation, an analytical study is carried out and a statistical approach to forecast the fragility of existing reinforced concrete buildings subjected to earthquake excitation is proposed. In the study, an ensemble of 50 existing RC buildings which represent the structural characteristics of existing structures, is selected. The building ensemble have experienced different levels of damages during the major earthquakes in Turkey, where the damage levels of the buildings can be specified as no-damage, slightly damaged, moderately damaged and heavily damaged. In the entire set, the number of stories of the buildings varies from 2-story to 8-story, therefore the set mostly consists of low- and mid-rise structures. All buildings with frame structural systems are investigated in details by means of their structural material quality, reinforcement amount, detailing of bars and local site conditions. Existing concrete class is found to be varying from fck=7~25.0 MPa, while the reinforcing steel is mostly S220 (fyk=220 MPa) class except 6 buildings that have S420 (fyk=420 MPa) bars. In order to conduct the analytical study, nonlinear dynamic and static analyses are carried out for this set of 50 frame structures utilizing the modified version of DRAIN-2DX (Prakash et al. 1993; Ascheim, 2005) computer program, which is capable of handling the stiffness and strength degradation through Takeda hysteretic model. Nonlinear behavior of columns and beams is assumed to be represented with a stiffness degrading hysteresis model. For the entire building stock, a structural damping of 5% and a strain-hardening of 3% are taken into account. The contribution of non-structural infill walls is also considered during the preparation of a number of 100 planar structural models for these buildings. The structural model for non-structural walls is established as in Al-Chaar and Lamb (2002), where walls are modeled by two diagonal compression struts. A set consisting of 7 artificial ground motions is generated with durations of 25 seconds which are compatible with the design spectrum defined in the Turkish Earthquake Resistant Design Code (TERDC-2007), that has a probability of exceedance of 10% within 50 years. Locations and local site conditions of the investigated moderately damaged buildings are also taken into account and the effective ground acceleration of A0=0.40g for seismic zone-1 and a local soil class of Z2 with characteristic periods of TA=0.15s and TB=0.40s are employed. This ensemble of artificial ground motions is employed during the analyses. In order to evaluate the acquired structural responses in an extended acceleration range, each of the artifical earthquake record is scaled from 0.1g up to 0.8g with an increment of 0.05g and therefore a set of 105 ground motion records is achieved. The demand parameters in terms of top story drift ratio and base shear demand ratio are obtained for the earthquake ensemble and after calculating the base shear capacities of each building in the stock, the demand/capacity ratios (DCR) are determined in terms of limit states. In the light of the analyses results, one of the main observations of this research is; existing buildings which reflect the general characteristics of existing structures have significantly high values of displacement demands due to their weaknesses in stiffness, however low values of base shear, depending on the poor concrete quality, insufficient amount of transverse reinforcement and their low structural strengths. Finally, evaluations are performed oriented to the damage potential of the existing buildings. A software is enhanced called Fragility Of REinforced Concrete Actual Structures in Turkey (FORECAST) in order to perform the statistical investigations to estimate the probable damage states of buildings, which reflect the general characteristics of existing structures with weakness in stiffness, strength and having poor concrete quality. By the favor of this software, damage levels are defined correlated with demand/capacity and interstory drift ratio parameters and fragility relationships are proposed. When a rapid estimation of the damage level of a building and/or building stock under the effect of a probable seismic event is necessitated, the proposed fragility relations are applicable for brittle and ductile structural systems by considering the related demand parameters for sub-standard existing buildings. Consequently, in order to investigate the congruity of the proposed functions along with the various researches focused on establishing fragility functions, comparisons are carried out in an attempt to designate the individuality and novelty that the dissertation reveals by means of structural responses.