It is generally accepted that the interaction between a bridge and its abutment＇s backfill soil is highly nonlinear, especially under a strong earthquake loading that contains a velocity pulse. For bridges with skew ...It is generally accepted that the interaction between a bridge and its abutment＇s backfill soil is highly nonlinear, especially under a strong earthquake loading that contains a velocity pulse. For bridges with skew abutments, the superstructure-abutment interaction can dominate the overall bridge performance. This study puts forth a new approach for predicting the lateral capacity of a skew abutment using verified high-fidelity three-dimensional continuum finite element （FE） models. The core idea is that the lateral capacity of a straight abutment is bounded from above and below by that of the abutment of a skew bridge that has the same deck-width, and that of another skew bridge （with the same angle） that has the same backwall length as the original/straight bridge, respectively. This postulation is then used in reverse to estimate the lateral capacity of a skew abutment, given the capacity of a straight but otherwise identical one with an arbitrary length. In prior research, the latter information had already been obtained in closed-form expressions that use physical parameters, such as backfill cohesion, internal friction angle and density, backwall height, and backwall-backfill friction angle. The approach presented here is constrained by the assumption that bridge deck will not rotate during loading. While this assumption is generally violated in a strong earthquake--because a skew bridge will tend to rotate, especially if its in-plane torsional rigidity is low, the model presented does serve as an anchor for parameterizing more advanced （e.g., macro-element plasticity） models that allow rotation, and also as fully parametric lateral response models for torsionally stiff （ile., multi-span, multi-bent） skew bridges.展开更多
文摘It is generally accepted that the interaction between a bridge and its abutment＇s backfill soil is highly nonlinear, especially under a strong earthquake loading that contains a velocity pulse. For bridges with skew abutments, the superstructure-abutment interaction can dominate the overall bridge performance. This study puts forth a new approach for predicting the lateral capacity of a skew abutment using verified high-fidelity three-dimensional continuum finite element （FE） models. The core idea is that the lateral capacity of a straight abutment is bounded from above and below by that of the abutment of a skew bridge that has the same deck-width, and that of another skew bridge （with the same angle） that has the same backwall length as the original/straight bridge, respectively. This postulation is then used in reverse to estimate the lateral capacity of a skew abutment, given the capacity of a straight but otherwise identical one with an arbitrary length. In prior research, the latter information had already been obtained in closed-form expressions that use physical parameters, such as backfill cohesion, internal friction angle and density, backwall height, and backwall-backfill friction angle. The approach presented here is constrained by the assumption that bridge deck will not rotate during loading. While this assumption is generally violated in a strong earthquake--because a skew bridge will tend to rotate, especially if its in-plane torsional rigidity is low, the model presented does serve as an anchor for parameterizing more advanced （e.g., macro-element plasticity） models that allow rotation, and also as fully parametric lateral response models for torsionally stiff （ile., multi-span, multi-bent） skew bridges.
