Pile supported river bridge failures are still observed in liquefiable soils after most major earthquakes. One of the recurring observations is the mid span collapse of bridges (due to pier failure) with decks falling into the river while the piers close to the abutment and the abutment itself remain stable. This paper proposes a mechanism of the observed collapse. It has been shown previously through experiments and analytically that the natural period of bridge piers increases as soil liquefies. Due to the natural riverbed profile (i.e. increasingly higher water depth towards the center of the river), the increase in natural period for the central piers is more as compared to the adjacent ones. Correspondingly, the displacement demand on the central pier also increases as soil progressively liquefies further promoting differential pier-cap displacements. If the pier-cap seating lengths for decks are inadequate, it may cause unseating of the decks leading to collapse. The collapse of Showa Bridge (1964 Niigata earthquake) is considered to demonstrate the mechanism. The study suggests that the bridge foundations need to be stiffened at the middle spans to reduce additional displacement demand.
Pile supported river bridges still continue to collapse after most major earthquakes in the event of liquefaction. The identified failure mechanisms of piles in liquefied soil are: bending failure due to the inertial loads of the superstructure and kinematic loads due to the lateral spreading of soil; shear failure due to shear loads; buckling instability failure due to vertical loads and associated imperfections; settlement failure due to loss of effective stress in the liquefied zone and finally failure due to the effects related to the elongation of natural period of the piers (also referred to as dynamic failure). This paper revisits the collapse of Shengli Bridge (due to 1976 Tangshan Earthquake) and Panshan Bridge (due to 1975 Haicheng Earthquake) based on the aforementioned failure mechanisms. It has been concluded that pile supported bridges in liquefiable soil can collapse due to each of these five failure mechanisms or due to a suitable combination thereof. It is therefore quite imperative to design pile foundations in liquefiable soil by taking all the failure mechanisms into consideration. The simplified calculation procedure presented in this paper can also be used to carry out the design of bridge piles in the liquefiable soil.
The behaviour of pile supported bridges in case of liquefaction during the earthquakes is not completely understood as can be seen from the failure of bridges during recent major earthquakes. It has been a recurring observation in most of the failure of pile supported bridges that the middle spans resting on the middle piers collapse, while the abutments and the piers close to them remain stable. Therefore, this study was carried out to investigate the mechanisms behind such midspan collapse of pile supported bridges in liquefiable soil deposits. Firstly, this thesis reports the simplified analytical expression developed to explain the midspan collapse of these bridges. It has been found that for the simply supported bridges, where each of the pier acts independently of the other, the natural period of the piers elongates in case of liquefaction due to increase in the unsupported length of the pile. Due to this process, the central piers of the bridge have higher natural period as compared to the adjacent ones, which in turn induces higher lateral displacement demand on the former. This phenomenon perpetuates differential lateral displacement for the adjacent piers. Hence, if enough seating length is not provided, the span may get unseated. Further, it has also been shown through the detailed case studies of collapse of around six bridges in various different earthquakes across the globe that this failure due to effects related to elongation of natural period of the piers can also make a bridge susceptible to fail in case of liquefaction, along with the other failure mechanisms. Further, it has been observed through the shake table tests that the natural frequency of the various pile supported piers of the bridge reduces during the course of liquefaction, with the central pier attaining the lowest natural frequency among all. Due to the increase in the flexibility of central pile owing to liquefaction, the maximum bending moment is observed at a shallower depth of pile, rather than at the interface of liquefied-nonliquefied soil. However, for the abutment piles, where the effect of lateral spreading is more as compared to any other piers, it has been found that the maximum bending moment is located at a section at the interface of liquefied-nonliquefied soil. Therefore, the design of piles for various bridge supports should be designed appropriately.