Experimental modelling of kinematic bending moments of piles in layered soils L. Dihoru, S. Bhattacharya, C.A. Taylor, D. Muir Wood University of Bristol, Bristol, United Kingdom F. Moccia, A. L. Simonelli University of Sannio, Benevento, Italy G. Mylonakis University of Patras ABSTRACT: The importance of scaled modeling in understanding seismic pile-soil superstructure interaction cannot be over-emphasised. The paper presents a series of 1-g shaking table tests aiming at clarifying certain aspects of kinematic and inertial interaction effects for specific problems. In the experiments a pile was placed in a layered deposit where the stiffness of the layers varied between two extremes [1.8 times and 80 times] and different earthquakes were fired. The piles were densely instrumented with strain gauges and therefore bending moments along the length could be inferred. Necessary similitude relations that need to be conserved in studying different aspects are derived. The importance of scaled input motion in understanding the kinematic behavior is explained. Certain broad conclusions are drawn. 1 INTRODUCTION Post-earthquake reconnaissance work (Mexico City (1985), Kobe (1995), Chi-chi (1999), Bhuj (2001), Niigata (2004)) has shown that a large number of pile-supported buildings built on loose to medium dense sands or layered soils suffered considerable settlement and tilting. In some of these cases (Mex- ico City 1985, Kobe 1995), damage has been ob- served along the piles, at the pile and/or close to in- terfaces separating soil layers with vastly different material moduli. The modern theories acknowledge that beside the inertial effects of the superstructure, the kinematic effects of the surrounding soil play a significant role on pile seismic behaviour. There is a rich body of analytical work in the area centred around continuum mechanics models. Notably, Winkler foundation models have been extensively used as they provide the ability to simulate non- linearity and stiffness degradation in the soil sur- rounding the pile (Penzien (1970), Matlock & Foo (1979), Novak (1979), Kavvadas & Gazettas (1993), Banerjee (1995), Mylonakis et al (1997), Tahghighi & Kanagai (2007)). In practice, these Winkler models are often based on non-linear “p-y models” where “p” denoted the pressure on the pile per unit length and “y” is the corresponding deflection. Readily available p-y curves for sand, soft clay and stiff clay for static loading are available in API (2003) and DnV (1992). Slight modifications on the static p-y curves to ac- count for cyclic loading under offshore conditions can be found in Bhattacharya et al (2006). These “p- y” models are empirical and the profession uses them mainly because of their successful application in the offshore industry for over four decades. Finite element and boundary element analyses have also been carried out in which the piles, the surrounding soils and the superstructure are repre- sented by axisymmetric elements and energy trans- mitting boundaries (Kuhlemeyer (1979), Krishnan et al (1983), Kaynia & Mahzooni(1996), Wu & Finn (1997), Zhang et al (2000), Padron et al (2008), Chau et el (2008)). With the emergence of increasingly sophisticated theories for soil-pile-superstructure interaction, it has become necessary to verify their validity via ex- perimental work. According to the size of the piles, the test medium and the technique employed, four types of experimental tests have been identified as productive: full-scale field tests, small prototype field tests, small scale laboratory tests (‘1-g tests’) and small scale centrifuge tests (‘n-g tests’). The de- tailed description of these categories is beyond the scope of this paper. A useful review of dynamic pile testing can be found, for example, in Novak (1987). The present work focuses on small scale model- ling of soil-pile-superstructure interaction in layered soils. The research was developed inside the RE- LUIS framework and represents the joint effort of