Abstract- Görk

ADVANCED MODELLING AND COLLAPSE RISK ESTIMATION OF A HIGHWAY BRIDGE

Serkan Görk

(Thesis Supervisor: Erdal Şafak & Thesis Co-supervisor: M. Nuray Aydınoğlu)

ABSTRACT

In this study, collapse risk of a highway bridge is estimated analytically using advance mathematical model. The concentration of this study is on advance modeling of the bridge, including linear sticks of beams, nonlinear modeling of piers and abutments with their shear capacities, inelastic connecting members, inelastic foundations and inelastic soil springs. The structural data including all geometry and dimensions, configuration of structural members and connecting elements, reinforcement details for piers, and material properties are taken from ‘as-built’ drawings of the bridge from contractors. Two different mathematical models have been constructed with different foundation conditions for nonlinear response history analyses with enough numbers of earthquake records. Collapse fragility curves are obtained in terms of spectral acceleration corresponding to dominant natural period of the bridge. The integration of collapse fragility curves of models and corresponding hazard curve gives the collapse risk of the bridge.

In the first chapter of this study, the scope of the work and general information about the bridge has been given. In the second chapter, the methodology of estimating collapse risk in an analytical way has been introduced. The steps of collapse risk estimation have been explained. In the third chapter, mathematical modeling of the structure has been given in details. Materials, nonlinear load-carrying members, beams, connecting elements, foundation and soil properties have been given in detail. In the fourth chapter, analysis objectives, modal analysis results, the earthquake records and hazard curves have been presented. In the fifth chapter, fragility curves are developed and the collapse risk has been calculated for these two models. In the last chapter, the conclusions are presented and suggestions for improving this study are proposed.