Parameterized fragility-based uncertainty influence quantification and sensitivity analysis methodology: Concept, formulation, and application

Abstract

Uncertainty modeling is a crucial issue in the seismic risk and resilience assessment of structures and infrastructures. Understanding and quantifying the impact of structural model parameter uncertainty (i.e., the epistemic uncertainty) on seismic fragility of structures are still open issues in the field of earthquake engineering. To this end, this study proposes a parameterized fragility-based methodology to quantify and rank the influence of multiple model parameter uncertainties on the seismic fragility result when their uncertainty is simultaneously involved. The concept and detailed procedure of this methodology is first introduced in this study. After that, the proposed methodology is applied to a case study bridge, i.e., unbonded laminated rubber bearing (ULRB) supported highway bridge retrofitted with transverse steel damper (TSD), to investigate how the uncertainty of the concerned model parameters affects the seismic fragility of the selected benchmark bridge when separately using the residual and peak displacements as engineering demand parameters (EDPs). The parameter sensitivity of fragility corresponding to these two EDPs is then quantified by a new proposed index and ranked using Tornado diagrams. In addition, the seismic fragility result when considering different sources of uncertainty is compared and discussed in this study through the parameterized fragility model. The result shows whether considering the uncertainty of model parameters has a more considerable influence on the fragility result when using the residual displacement as EDP as compared with using peak displacement as EDP. The sensitivity ranking of the parameters is affected by the damage level or the nonlinearity of the structural system. Regarding the residual and peak bearing displacements of ULRB-supported highway bridges, the friction coefficient is the most sensitive parameter.