Liquefaction-induced lateral ground displacement has caused significant damage to pile foundations during past earthquakes. Ground displacements due to liquefaction can impose large forces on the overlying structure and large bending moments in the laterally displaced piles. Pile foundations, however, can be designed to withstand the displacement and forces induced by lateral ground displacement. Piles may actually “pin” the upper layer of soil that would normally spread atop the liquefied layer be low it into the stronger soils be low the liquefiable soil layer. This phenomenon is known as the “pile-pinning” effect. Piles
have been designed as “pins” across liquefiable layers in a number of projects, and this design methodology was standardized in the U.S. bridge design guidance document MCEER/ATC-49-1. A number of simplifying assumptions were made in developing this design procedure, and several of these assumptions warrant re-evaluation. In this report, some of the key assumptions involved in evaluating the pile-pinning effect are critiqued, and a simplified probabilistic design framework is proposed for evaluating the effects of liquefaction-induced displacement on pile foundations of bridge structures. Primary sources of uncertainty are incorporated in the proposed procedure so that it is compatible with the Pacific Earthquake Engineering Research (PEER) Center performance- based earthquake engineering (PBEE) framework.
A detailed description of the problem and of the current approach to evaluating liquefaction-induced bridge damage is first presented. The PEER-PBEE methodology is described, with emphasis on the components of the method that are more pertinent to the problem under study. Several preliminary
evaluations are performed before the proposed simplified procedure is applied. These preliminary assessments include the estimation of the seismic hazard at the site, a liquefaction triggering assessment, and an evaluation of the liquefaction-induced flow failure potential at the site.
The details of the application of the PEER -PBEE methodology to the case of bridges founded on piles affected by liquefaction-induced lateral ground displacement are then provided. A procedure to estimate residual lateral displacements at the abutments for a given intensity of the ground motion is described, followed by a discussion of the structural models that can be used to evaluate how these displacements influence the response of the bridge superstructure. With these assessments of bridge response, which are couched in probabilistic terms with key sources of uncertainty characterized,
well-defined damage states of the bridge system and its components can be estimated. Lastly, procedures for the final estimates of repair costs and downtimes for the most likely damage conditions are presented.
The proposed procedure is applied to a realistic bridge example that was developed by PEER and Caltrans engineers to illustrate the use of the method and the insights that can be garnered from its application to a bridge evaluation at a site with potentially liquefiable soil. Three alternative models are available in the estimation of repair cost ratios and downtimes for the different hazard levels included in the analysis. They are the first-order reliability method, the point estimate method, and the simplified coupled model. Some of the most useful insights developed through the application of the proposed method are the identification and quantification of the parameters most affecting the uncertainty in the assessment of the seismic performance of a bridge. The ground motion hazard, which was defined by the spectral acceleration at the degraded period of the potential sliding mass at the abutments, is the most important source of uncertainty. It is followed by the uncertainty in estimating seismic displacement at a specific ground motion hazard level and the uncertainty in estimating the residual undrained shear strength of the liquefied soil.
The proposed simplified procedure is validated through its application to three important “case” histories: Landing Road bridge (1987 Edgecumbe earthquake), Showa bridge (1964 Niigata earthquake), and one centrifuge model test performed at UC Davis. The results from this procedure compare well with the observations from these cases, and they are also consistent with the results from an advanced finite element analyses performed by the University of Washington research team. Lastly, a procedure to incorpor ate ground motion time histories is described, and findings from the study and recommendations for future work are summarized.
A simplified “user’s guide” that summarizes the proposed procedure follows this abstract.
Full List of PEER Reports: click here.