Local engineering demand parameters for seismic risk evaluation of low ductility reinforced concrete buildings

The damage occurred during recent earthquakes in many existing reinforced concrete (RC) buildings designed before the introduction of modern anti-seismic codes has shown that these structures are very vulnerable to the seismic action due to their reduced ductility capacity. This underlines the need to develop retrofit techniques for reducing the vulnerability of existing structures and of reliable tools for assessing the effectiveness of the retrofit and the resulting structural safety. Earthquake risk mitigation and seismic risk evaluation are among the engineering's most complex challenges. Seismic risk evaluation must include a proper assessment of the system vulnerability and seismic hazard. Seismic vulnerability analysis of structural systems can be a rather hard task especially, amongst others, due to the high degree of uncertainty affecting the system properties, the capacity and the input definition and hence should be performed within a probabilistic framework. The Performance Based Earthquake Engineering (PBEE) framework introduced by the Pacific Earthquake Engineering Research Centre (PEER) is currently the more advanced probabilistic methodology adopted in seismic engineering. The effort of many researchers has been employed along the last decades to reach the present state of development, however, there are still drawbacks in the design and seismic performance assessment of certain type of structures. The objective of this Thesis is to investigate the probabilistic response and vulnerability of a class of low ductility reinforced concrete (RC) frame buildings and investigates about the use of buckling retrained braces as retrofit technique. Traditionally, structural response is measured by Engineering Demand Parameters (EDPs), such as the overall maximum interstory drift over the height of the building. The use of this EDP is adequate to describe the seismic response of ductile frame designed by strength hierarchy rules, but may lead to a high approximation in the vulnerability evaluation since in this case there is not direct relation between local failure mechanism and global interstory drifts. To obtain a more thorough characterization of the vulnerability of the structure, a multi-component fragility study is necessary, hence, the use of local EDP is investigated in order to enable a more realistic and thorough description of the failure mechanisms for structural vulnerability. The study proposes an optimized methodology for the probabilistic evaluation of seismic demand of low ductility RC frames by exploring a range of intermediate, local and global EDPs, identifying appropriate regression models and comparing performances of different ground motion intensity measures used in the probabilistic analysis. In particular, different EDPs are considered in order to highlight the most significant failure modalities in RC lowductility frame buildings, optimal PSDMs of single components are developed for various EDPs, and the viability of alternative IMs is explored. While the form of PSDMs and efficiency of various IMs has been readily explored for global response parameters of RC buildings, this study provides insight into the form of regression model appropriate for such local and intermediate level EDPs as steel and concrete strains, moments and shears on beams and columns, and global responses such as base shear or story accelerations. Furthermore, ground motion IMs are analyzed to identify the IMs that are most appropriate for Probabilistic Seismic Demand Analysis (PSDA) of low ductility RC frames on the basis of such characteristics as IM efficiency and sufficiency. Additionally, the uncertainty about these demand models is assessed including hypothesis tests of the typical lognormal distribution of demands and homoscedasticity assumptions. All the considerations are based on the results of a PSDA performed on a case study. Moreover, a probabilistic methodology for assessing the vulnerability of existing RC buildings with limited ductility capacity and retrofitted by means of dissipative braces is proposed. The methodology use local EDP in order to develop component and system fragility curves of the bare and the retrofitted frame and allows to evaluate the safety level reached by the frame before and after the retrofit by taking into account the probabilistic properties of the seismic response and at the same time, employing an efficient structure dependent IM. The proposed approach allows to highlight the possible changes in the most significant collapse modalities before and after the retrofit and to evaluate the effectiveness of the retrofit. A benchmark 2-dimensional reinforced concrete frame with low ductility capacity is considered as case study. The frame is designed for gravity-loads only and does not comply with modern anti-seismic code requirements. It is retrofitted by introducing elasto-plastic dissipative braces designed for different levels of their target base-shear capacity, following a design method involving the pushover analysis of the bare frame. The obtained results show the effectiveness of the use of component level vulnerability evaluation of low ductility frames, and the effectiveness of the methodology in describing the changes in the performance due to retrofit. The lack of accuracy as consequence of the use of these global EDPs for evaluating vulnerability of retrofitted frames is posed in evidence. Finally, an increase of the dispersion of the retrofitted frames with respect to the bare frame is shows and this result implies the importance of considering the dispersion in the evaluation of seismic safety level achieved after the retrofit. The proposed methodology also allows testing the effectiveness of this simplified criterion employed for the design of braces.