ABSTRACT
Furthermore, recent centrifuge and full-scale 1g shaking table tests of single piles and pile groups indicate that the permeability of the liquefied sand is an extremely important and poorly understood factor, with a suggestion that the pile bending moments in silty sand may also be much greater than in clean sand. This is illustrated by two centrifuge tests of a 2 × 2 pile group (Figure 1) conducted at Rensselaer Polytechnic Institute (RPI) by Gonzalez (2005), using the same fine sand, but with water as pore fluid in one of the tests, and a viscous pore fluid in the other test, hence simulating two sands of widely different permeabilities in the field. A flexible model container (laminar box) was used in these and other lateral spreading tests, in order to simulate the shear beam free field conditions and to allow development of the
1 INTRODUCTION
Liquefaction-induced lateral spreading of sloping ground and ground near waterfronts continues to be a major cause of damage to deep foundations. In the US, Japan and other countries, buildings, bridges, and other structures supported by deep foundations have been damaged in many earthquakes, with billions of dollars in damages. The observed damage and cracking to piles is often concentrated at the upper and lower boundaries of the liquefied soil layer where there is a sudden change in soil properties; or at the connection with the pile cap. Case histories, as well as 1g shaking table and centrifuge model tests, indicate that the effect of lateral spreading on piles can be characterized in first approximation as a pseudostatic, kinematic soil-structure interaction phenomenon, driven by the permanent lateral movement of the ground in the free field. Even though various foundation analysis and design methods have been proposed, where the soil applies static lateral forces to the pile foundation, there is currently a huge uncertainty associated with the maximum lateral pressures and forces applied by the liquefied soil, which translates into a similar huge uncertainty in the calculated maximum pile bending moments. For example, in the Japan Road Association (JRA, 1996) method, the lateral pressure is specified as 30% of the total overburden pressure, while Abdoun et al. (2003)
lateral spreading. Figure 1 shows the measured bending moment at 6 m depth in the same pile for the two tests, versus pile cap displacement. In the “water” test, the pile head reached a maximum prototype displacement of 7.5 cm and maximum bending moment of 62 kN-m and then bounced back, while in the “viscous fluid” test, the pile head reached a maximum displacement of 42 cm and a maximum bending moment of 425 kN-m at the end of shaking, without bouncing back. This is a factor of 425/62 = 7 between maximum pile bending moments. Therefore, the uncertainty in lateral soil forces and pile bending moments, related to our poor understanding of the complex behavior of liquefied soils in the vicinity of foundations, can produce maximum lateral liquefied soil forces and pile bending moments varying by factors as high as 3 or 7. This certainly constitutes a critical “gap” in our current earthquake engineering knowledge.
