The main goals of this research were to evaluate typical 1950s and 1960s as-built bridge columns in western Washington State in large subduction zone earthquakes and to investigate the dependency of failure mechanisms on loading history. Eight displacement histories were applied to eight nominally identical, half-scale, circular reinforced concrete columns expected to respond primarily in flexure (flexure-dominated). The main design deficiencies were a short longitudinal reinforcement lap splice at the base of the column (35d sub b) and inadequate transverse reinforcement. Test results showed that the failure mode of reinforced concrete columns was controlled by the column loading history. Three distinct failure mechanisms were observed for columns with an aspect ratio of approximately 4.2, assuming symmetric, double-curvature behavior. Large initial displacements greater than six times the effective yield displacement (+öy) were likely to result in shear failures. Columns experiencing many displacements less than 4+öy were likely to fail because of longitudinal reinforcement buckling. Columns subjected to several displacement excursions less than 4+öy followed by an excursion greater than 6+öy were likely to fail by longitudinal reinforcement slipping within the splice region. Despite the deficiencies present in circular reinforced concrete bridge columns built before 1975 in western Washington State, this study showed that flexure-dominated columns with a 35d sub b lap splice in multiple-column bent, three- or four-span bridges were not likely to experience significant damage in the predicted Cascadia Subduction Zone earthquake. However, other components of the bridge need to be assessed to determine whether the global bridge response is acceptable under the predicted Cascadia Subduction Zone earthquake.
Abstract