Unseating mechanism of a skew bridge with seat-type abutments and a Simplified Method for estimating its support length requirement.

• Unseating mechanism of a skew bridge is investigated by experiments. • An existed mechanism is modified based on shake table experiment results. • A Simplified Method is developed to estimate the support length of skew bridges. • The Simplified Method is validated by shake table experiment results. • The SM may be used for the preliminary design of bridges with standard gap sizes. In this paper, the unseating mechanism of a skewed bridge during earthquake shaking is investigated. To begin, an existed unseating mechanism in literature, which was proposed by other researchers, is described. The mechanism hypothesizes that under lateral loading, the obtuse corner of the superstructure of a skew bridge engages the adjacent back wall and the superstructure then rotates about this corner, causing large in-plane displacements at the acute corner at the other end of the span, potentially unseating the edge girder. It is then validated using data from a comprehensive set of shake table experiments recently conducted by the author and other researchers. Whereas the experimental data confirms this mechanism as the primary reason for the rotation, it is not the only reason. Impact of the span against the back wall is followed immediately by rebound away from the wall and the span continues to rotate in the same direction in free vibration about the center of stiffness of the substructure. This additional rotation further increases the superstructure rotation and the in-plane displacement of the acute corner. Based on the existed mechanism, a Simplified Method is then developed to estimate the support length requirements of skew bridges. This method, which is based on response spectrum analysis, can consider the closure of expansion gap. The basic procedures are described for the AASHTO's design spectrum, and spectrum of an arbitrary ground motion. A closed form solution is achievable for the former spectrum, while iteration is required for the latter spectrum. The results are then compared with the experimental dataset. It is found that the accuracy of the Simplified Method depends on the gap size. That is: it overestimates the maximum displacements normal to the abutment of skew bridges when the gap is small, accurately estimates when the gap is in a typical range, and underestimates when the gap is large. The underestimation is mainly because the Simplified Method neglects the "bounce back" effect of skew bridges. Therefore, the Simplified Method may be used for the preliminary design of bridges with standard gap sizes, but should be modified to include "bounce back" effect when the gap is large. [ABSTRACT FROM AUTHOR]