5. Conclusions
(1) The authors conducted seismic analyses to simulate the distinctive damage process of pile foundations where the characteristics of pile damage were observed to vary depending on pile position even in the same building and the same building frame, during the 2011 off the Pacific coast of Tohoku earthquake. The analysis results provide a reasonable explanation of the observed damage. The analysis model considered for the foundation structure provides a reliable evaluation of nonlinear behavior of the pile-soil system. The pile group effect is also taken into account depending on the pile position in relation to the loading direction.
(2) The analysis method considered 2 phases of the damage process to provide an efficient trace of the observed damage to piles. In Phase 1, the axial loads in Piles No. 1 and 2 increased under north-to-south loading, and the shares of horizontal load also increased, resulting in compression or shear failure at the pile head. In Phase 2, the remaining piles were subjected to pull-up forces under the sequence of the west-to-east loading, reduction in the rotational stiffness at the pile head, and damage to the middle part of the piles.
(3) There is high possibility that partial and consecutive damage occurs during severe earthquakes, particularly when building has an irregular shape and/or unequal load distribution. Design based on simplified method that assumes elastic behavior of piles, fixed conditions at the pile head, and constant soil springs, is completely inadequate in representing the pile damage mechanism. It is necessary to consider nonlinear load-deformation characteristics of piles, to construct an adequate model of the pile head connection with respect to axial load and to take into account the nonlinear behavior of soil springs including pile group effects for seismic design pile foundations, particularly while considering severe earthquake shaking. It should be noted that the influence of axial load is remarkable.
