Behavior of liquefying sand and CISS piles during full-scale lateral load tests

Recent earthquakes have increased awareness for the need to increase understanding of liquefied soil behavior in response to laterally loaded pile foundations. Historically, small-scale experiments have been conducted to investigate the behavior of soil and piles during the liquefaction process. While these tests do provide insight into soil and structural behavior, the quantitative results have not gained widespread acceptance among engineering practitioners.; To address concerns with small scale model tests and to provide increased understanding concerning the behavior of liquefied soils, a series of full-scale lateral load tests were conducted on a 0.6-m and 0.9-m diameter Cast-In-Steel-Shell (CISS) piles. Data obtained from the instrumented pile foundations has been used to back-calculate p-y curves for sand at various excess pore water pressure ratios.; The back-calculated p-y curves from the full-scale tests are compared to p-y curves back-calculated from a centrifuge model test and to p-y curves typically used in the design of pile foundations in liquefied sand. The p-y curves back-calculated from the full-scale and centrifuge tests are similar in shape and magnitude of resistance. However, the full-scale p-y curves show more displacement is required to provide the same level of resistance. The full-scale p-y curves appear to provide a lower bound response for liquefied soils.; The comparison between typical p-y curves used for design of pile foundations in liquefied soil with those back-calculated from the full-scale tests show that the typical p-y curves provide too much resistance at small displacements and not enough resistance at large displacements. The use of typical p-y curves should be used with caution, having an understanding of the impacts these p-y curves will have on the design of the piles and superstructure.