Seismic Performance of a Rocking Pile Group Supporting a Bridge Pier

Abstract

Recent research has shown that full mobilization of foundation bearing capacity may be beneficial in terms of structural integrity—especially in the case of seismic motions that exceed the design limits. Full mobilization of foundation bearing capacity may serve as seismic isolation because it limits the inertia loading transmitted to the superstructure. Although most research has focused on rocking shallow or embedded foundations, a rocking pile group has attracted much less attention. A potential reason is the probability of structural damage below the ground level (at the piles), which may be difficult to repair or even detect. To shed more light on the problem, the present study investigates the seismic performance of a rocking pile group in clay, aiming to assess its efficiency as a seismic isolation alternative. Employing the finite-element (FE) method, an idealized yet realistic example of a single bridge pier supported by a pile group foundation is analyzed. A carefully calibrated and thoroughly validated kinematic hardening constitutive model is employed for the soil, and the concrete damage plasticity model is applied for the structural members. Using a suite of records as seismic excitation, the response of an intentionally underdesigned rocking pile group is compared with that of a conventionally (capacity) designed system. Similarly to what has been shown for shallow foundations, the comparison reveals that the rocking pile group can be beneficial for the seismic performance of the bridge, reducing the flexural demand on the pier at the expense of increased settlement. Interestingly, the rocking pile group exhibits a genuinely ductile response, such that none of the studied ground motions could lead to full mobilization of the bending moment capacity of the piles. Thus, pile structural damage is avoided. The findings of the present study reveal the advantages of exploiting nonlinear soil–foundation response and indicate that there is a great potential to optimize the contemporary seismic foundation design, which conventionally culminates in massive pile group foundations. The rocking pile group concept may be of particular interest for the retrofit of existing bridges that do not meet the requirements of the current seismic design provision because it can reduce or even completely avoid strengthening the foundation. Ultimately the presented findings call for a shift toward performance-based design, with due consideration of geotechnical failure modes.