Shake Table Studies on Soil-Abutment-Structure Interaction in Skewed Bridge

Soil-abutment-structure interaction could affect the seismic response of bridges considerably. Skew angle might significantly influence the mobilized passive resistance of the backfill soil and the behavior of soil-abutment system due to the large induced in-plane rotations and translation of the superstructure, coupled with variations in stiffness and strength of backfill soil in skewed abutments. The current Seismic Design Criteria of the California Department of Transportation (Caltrans) does not include any special consideration for the skew angle effect on the passive capacity of soil-abutment systems. Previous experiments on skewed abutments were undertaken on abutments that were restrained against rotation with prescribed uniform displacements tested by gradually increasing lateral loads under static conditions, with no dynamic effect simulated. The effects of abutment rotation, impact on the abutment wall and dynamic earthquake forces were not studied.;The overall objective of the current study was to investigate experimentally and analytically the effect of skew angle on the abutment soil response under realistic dynamic earthquake loading and develop recommendations on modeling of skewed abutments for application in bridge seismic design.;The experimental study was focused on soil-abutment-structure interaction in skewed bridges under dynamic loading based on large-scale shake table tests at the University of Nevada, Reno. Three 5.5-ft high abutment walls at three skew angles of 0º, 30º, and 45º with a projected width of 10 ft in the direction of motion were impacted by a bridge superstructure and pushed in the longitudinal direction of the bridge into a 25 ft long by 19 ft wide engineered backfill soil embankment in a stationary timber box. The abutment walls were allowed to rotate to further simulate actual bridge abutments realistically. The bridge superstructure was represented by a concrete bridge block supported on elastomeric bearings that simulated the stiffness of the substructure and a mass that accounted for similitude effect. The skew angle of the bridge block was changed in different experiments to match the angle of the abutment wall. The bridge block was supported on a shake table. The 1994 Northridge Sylmar earthquake record was simulated in the table with successively increasing amplitudes. A large number of transducers of different types were used to monitor the bridge block and the abutment response under the simulated lateral dynamic loading. The experiments verified that skewed bridges tend to rotate in the direction of reducing the skew angle. This corresponds to impacting abutment at the obtuse corner and unseating of superstructure at the acute corner. The test results showed that the passive capacity, heaves, and accelerations of soil were reduced by increasing the skew angle although the abutment wall width increased when a higher skew was simulated. The distribution of backfill pressure across the abutment was primarily dependent on direction of the abutment wall rotation while the maximum pressure, heaves and accelerations occurred at the obtuse corner of the bridge block.;Analytical studies were performed by developing FLAC3D models of the shake table tests in the current study. The analytical models simulated the abutment wall and backfill under the static uniform and non-uniform displacement loading on the wall. Results from the analytical studies indicated that the backfill passive capacity was reduced when the abutment rotation was accounted for. The displacement contours from the analytical models that simulated the abutment wall rotation were similar to those obtained in the shake table tests. Design recommendations were developed by evaluating the most recent available models estimating the passive force-displacement relationships of the abutments considering the effect of skew.