Probabilistic Seimic Fragility And Risk Analysis Of Reinforced Concrete Frame Structures

In recent years, the new generation of performance-based earthquakeengineering (PBEE) aimed at controlling seismic risk and earthquake loss hasbecome one of world-wide research hot topics in the community of earthquakeengineering. Due to the strong randomness and uncertainties in earthquake srongmotions and engineering structures, the Pacific Earthquake Engineering ResearchCenter (PEER) proposed a fully probabilistic risk-informed decision framework forthe new generation of PBEE. In this PEER’s PBEE framework, probabilistic seismicrisk is defined as the seismic performance objective, which includes three mainbuilding blocks: probabilistic seismic hazard, probabilistic seismic fragility, andprobabilistic seismic loss.To adapt to the development trend of PBEE, a pilot edition of “general rule forperformance-based seismic design of buildings” was firstly published in2004inChina. Moreover, the new edition of the Chinese Seismic Design Code of Buildings(GB50010-2010) also carried out a performance-based seismic design method at thelevels of structural elements. However, both of the above important codes employedthe first generation of performance-based seismic design methodology. In fact, thenew generation of fully probabilistic PBEE has not been widely studied and appliedin China. In addition, the great Wenchuan earthquake in2008indicates us that eventhough the structures are designed according to the current seismic code, they arestill probable to collapse in face of unpredicted rare earthquakes or even “hugeearthquakes”. Therefore, it’s reasonable to evaluate seismic risk and safety ofcode-conforming structures. The assessment of seismic risk and safety for structurescan help make rational decisions on seismic retrofitting of structures and disastermitigation planning. Based on the above considerations, in this thesis, theassessment of probabilistic seismic fragility and risk for Chinese code-conformingreinforced concrete (RC) frame structures in the framework of new-generationPBEE is taken as research objective. From the viewpoint of uncertainty propagation,the analytical functions are derived for probabilistic seismic fragilty and risk. Usingthe derived analytical functions and highly effective simulation techniques, theprobabilities of seismic fragility, risk, and safety of Chinese code-conforming RCframe structures are systematically evaluated. The main contents are summarized asfollows:1) Twenty-three RC frame buildings with different fortification intensities andstoreys are designed according to Chinese codes as the “index archetype structures”. These structures are modeled in the platform OpenSees. Through comparing withthe test data of RC elements (columns) and a global structure, the OpenSees modelsare verified and validated to be adequate to describe the nonlinear behavior of theindex archetype structures.2) An advanced point estimate method incorportating marginal distributionsand correlation information of basic random variables is put forward by combiningdimension-reduction integration with Nataf transformation. In the presentedapproach, a new sensitivity factor is proposed. Through three numerical examples, itis illustrated that the developed method has adequate accuracy to estimate probilitymoments of model outputs. Moreover, the presented sensitivity factor is valid toanalyze the influences of the variation of random variables in the standard normalspace on the outputs of nonlinear complex models.3) One hundred real ground motion records are selected as inputs toprobabilistically evaluate the earthquake intensity measures through relating60intensity measures vs.6structural response parameters. The probabilistic seismicdemand models (PSDMs) for individual buildings are built by both cloud-methodand stripe-method, and the rationality of the log-linearity relationships betweenseismic demand and earthquake intensity is furhter verified. Since the conventionalPSDM cannot incorporate the case of collapse, a modified PSDM consideringcollapse case is presented. A new cloud-stripe method is developed for probabilisticseismic demand analysis for group buidlings, and the PSDM for group structures isbuilt up by the new approach.4) The first three non-collapse limit states (i.e. minor damage, moderatedamage, and severe damage) are identified from the pushover curves of structureswith a local-global hybrid damage controlling criterion. Considering the effects ofstructural random parameters on seismic capacity of buildings, a random pushoveranalysis approach with the advanced point estimate method is developed forprobabilistic seismic capacity analysis (PSCA) of non-collapse limit states ofstructures. To consider the effects of both randomness in structural parameters anduncertainty in earthquake strong motions on seismic collapse capacity of structures,a random incremental dynamic analysis (IDA) approach based on mean value firstorther second moment (MVFOSM) is developed for PSCA of collapse state ofstructures. The probabilistic seismic capacity models for23index archetypebuildings are built by the random pushover analysis approach and the random IDAmethod. Then, the probabilistic seismic capacity models for group structures arefitted from the PSCA results for all the index archetype structures. In addition, thethreshold values for the four limit states determined in this study areprobabilistically evaluated.5) The consistence between the intensity-based and displacement-based seismic fragility functions is proved. From the viewpoint of uncertainty propagation,the analytical formulations for probabilistic seismic demand and damge fragilityfunctions considering aleatory uncertainty are derived, and the point and intervalestimate functions of seismic fragility considering epistemic uncetianty are furthergiven. Then the interval confidence level of the point estimate function of seismicfragility is obtained. Using these analytical fragility functions, the seismic fragilitycurves for the index archetype structures are generated. Then the failureprobabilities of various limit states and damage states are calculated for all thearchetype structures from the obtained seismic fragility curves. A new index tomeasure structural safety named as “safety marginal ratio (SMR)” is proposed toevaluate the safety margins of the index archetype structures at various limit states.The HAZUS-compatible seismic fragility curves are generated, and compared withthe recommended seismic fragility curves in HAZUS. Considering the case ofcollapse, the conventional seismic fragility functions are modified. Seismic fragilityanalysis is further carried out for group structures, and then the generated seismicfragility curves for group structures are compared with those for each indexarchetype building.6) Incorporationg the generally used power-law seismic hazard function, theprobabilistic seismic risk functions considering both aleatory and epistemicuncetianties are derived based on the developed analytical fragility functions. Usingthese analytical functions, the probabilistic seismic risks of the index archetypestructures are evaluated at the levels of both seismic demand and seismic damage,based on the results of probabilistic seismic fragility analysis. To evaluate theprobabilistic safety of Chinese code-conforming RC frame structures, the failureprobabilities during the service life (50years) are further computed for the indexarchetype structures. Based on the calculated probabilistic seismic fragilities ofgroup structures, the probabilistic seismic risks for group structures are furtherevaluated.