Failure behaviors of in-situ deteriorated reinforced concrete T-girder bridges under cyclic loading is experimentally observed and a finite element modeling technique to predict the behaviors is developed. In this study, full-scale destructive tests of in-situ bridges are performed by applying cyclic loads up to failure, and modeling techniques for the non-linear finite element analysis of in-situ deteriorated reinforced concrete T-girder bridge are presented. Two in-situ reinforced concrete T-girder bridges were selected for the failure tests and the analysis, one a symmetrically loaded bridge and the other a non-symmetrically loaded bridge. Path-dependent in-plane constitutive laws of cracked reinforced concrete were utilized for material modeling of the analysis. An RC zoning method was applied to two-dimensional finite element modeling of the symmetrically loaded bridge and a combination of frame elements utilizing the fiber technique and layered shell elements were used for three-dimensional modeling of the non-symmetrically loaded bridge. Experimental results indicate that significant load carrying capacity is retained in old reinforced concrete bridges and analysis results show that the manner of modeling of degraded support conditions significantly affects the predicted responses of the capacity as well as the stiffness of the bridges. This significant effect of support conditions of the deteriorated RC bridges is verified and a simple modeling technique for the support condition is proposed to consider the degradation of supports. By applying the proposed modeling to the boundary condition of the bridges, a finite element failure analysis is carried out for the bridges subjected to cyclic loading. Then, the analytical results are compared with full-scale failure test results. The comparison shows that the finite element analysis technique along with the proposed boundary condition can be effectively applied to the failure analysis of in-situ reinforced concrete T-girder bridges subjected to cyclic loading.
Read full abstract