(e.g.,Mononobe and Matsuo 1929, Sherif et al. 1982, Bolton and Steedman 1982, Steedman and Zeng 1991, Stadler 1996). Many of these tests were performed on a very small scale, and the results have shown varying levels of agreement with the theoretical predictions (Al Atik and Sitar 2008). Koseki et al. (1998) suggested modifications to the M-O method based on experimental results from shake table excitation and tilting tests performed with a 0.5 meter tall wall. These modifications consider the possibility of a weakened band of backfill material (strain localization) existing along a previously formed active failure wedge. The experimental and analytical results of Koseki et al. (1998) suggest that the plane formed by that initial wedge might control the consecutive mobilization of earth pressure until shaking levels are strong enough to form a new, larger wedge in the stronger surrounding backfill. Deewoolkar et al. (2001) performed centrifuge dynamic excitation tests withfixed-base cantilever walls supporting saturated, liquefiable, cohesionless backfills.From those experiments, Deewoolkar et al. (2001) concluded that excess pore pressure generation contributed significantly to seismic lateral earth pressure in the saturated backfill. Deewoolkar et al. (2001) also found that the maximum dynamic thrust was proportional to the input base acceleration. Ling et al. (2005) conducted shake table tests on 2.8 meter high modular-block geosynthetic-reinforced soil walls. From those large scale tests, Ling et al. (2005) found that existing design methods underestimate the capacity of such flexible systems. Nakumara (2006) and Al Atik and Sitar (2008) recently conducted separate shake table tests using centrifuge facilities, and both separately concluded that the measured earth pressure during shaking was lower than the M-O method predictions.
Nakamura (2006) also found that the inertial force was not always transmitted to the wall and backfill simultaneously.1.2.4 Numerical StudiesSimulations of the dynamic wall-backfill interaction using numerical models haveprovided additional valuable insights. Alampalli and Elgamal (1990) developed anumerical model based on the compatibility between mode shapes of the wall and theadjacent backfill soil. Using a model consisting of flexible cantilever wall supporting a semi-infinite uniform visco-elastic layer, Veletsos and Younan (1997) concluded that the magnitude and distribution of wall displacement and pressure can be quite sensitive to the flexibilities of the wall and its base. Richards et al. (1999) presented a kinematic model with springs representing the soil and found that the point of action of the dynamic earth pressure resultant varies with different types of wall movement.Gazetas et al. (2004) performed
simulations of L-shaped walls, prestressedanchoredpile walls, and reinforced soil walls, employing both linear and non-linear soilmodels. Using those models, Gazetas et al. (2004) showed that including realistic effects such as the wall flexibility, foundation soil deformability, material soil yielding and soilwall separation and sliding tends to reduce the effects of dynamic excitations on those walls. Gazetas et al. (2004) also used an FE model to simulate a case history in which a retaining wall performed well during an actual earthquake.Psarropoulos et al. (2005) performed FE simulations of rigid and flexible nonslidingwalls and found that the rigid wall case converged to the Wood (1973) analyticalsolution, and the flexible wall matched the Veletsos and Younan (1997) solution. After achieving these results, Psarropoulos et al. (2005) extended their model to find that adding inhomogeneous soil layers further complicated the response, suggesting that simply adding a rocking spring to the model base may not accurately account for the wave propagation effects from the underlying foundation layer.Jung and Bobet (2008) added a translational spring to the base of a bending androtating wall model supporting elastic soil elements, and found that the wall rotational, bending, and translational flexibilities significantly affected the magnitude anddistribution of the dynamic pressure. Specifically, Jung and Bobet (2008) found that the dynamic earth pressure behind a rigid wall with a stiff foundation is larger than that for a flexible wall with a soft foundation.1.2.5 Dynamic Earth Pressure SummaryThe limit equilibrium Mononobe-Okabe (1926, 1929) equations for yielding wallsand the elastic soil-rigid wall solution of Wood (1973) provide methods for predicting the dynamic earth pressure. However, the observed field retaining wall seismic performance, shake table tests, and numerical investigations have not consistently agreed with those predictions. Rather, they have demonstrated that the complexity of the dynamic retainingwall-backfill response may require more detailed analysis and further experimentation in order to aid in refining design standards.1.3 Objectives and ScopeIn bridge seismic design, accurate modeling of the passive earth pressure forcedisplacement resistance at the abutments and pile caps leads to safer and more economic design. However the few available bilinear design models (e.g., Caltrans 2004, AASHTO 2007) and limited data from large scale tests, do not account for the variation in nonlinear passive force-displacement response that occurs in the field. In addition, the existing test data (from experiments performed on static backfill) and design models do not account for potential inertial effects on the structure and backfill which may influence the actual available resistance during shaking. Significant uncertainty also remains in the evaluation of dynamic earth pressure for retaining wall seismic design. The performance of many walls during earthquakes is not consistent with the predicted response based on the existing theory. Furthermore, very few well- documented case histories and large scale test data are available for validation of improvements to prediction methods. An excellent opportunity to provide new experimental insight into passive and dynamic earth pressure was provided as a component of a collaborative NEES investigation into the seismic performance of highway bridges (Saiidi 2004). As the primary component of that study, Saiidi (2004) performed ¼ scale 4 span bridge tests on 3 movable shake tables (Figures 1.14 and 1.15) at the University of Nevada at Reno (UNR). Within this collaborative framework, an abutment investigation was also designated to be performed at UCSD, using a large soil container on the outdoor shake table at the Englekirk Structural Engineering Center (ESEC). The work presented in this dissertation was conceived to maximize this provided abutment investigation research opportunity, by employing a single test configuration, along with numerical models, in order to address the following objectives:1) Investigate the passive earth pressure force-displacement relationship. First, performexperiments to record this relationship behind a model abutment wall, full scale inheight, with dense sand backfill. Next, test our ability to predict the peak passiveresistance and the force-displacement relationship using soil strength and stiffnessparameters determined from laboratory tests. Finally, use calibrated finite element(FE) model simulations to compare and provide curves for a range of backfill soiltypes and wall heights.2) Develop a model which can represent the backfill passive resistance at the abutmentsusing the provided experimental and numerical curves in dynamic bridge simulations.3) Record dynamic earth pressure on the vertical wall by subjecting the testconfiguration to shake table excitation. Provide new insight into the recorded levelsof earth pressure during shaking, including any interesting effects observed throughthe experimental data. Test our ability to predict the dynamic earth pressure usingexisting theory.4) Develop FE models which can provide reasonable estimates of the recorded dynamic earth pressure, and capture key aspects of the wall-backfill response.5) Record dynamic passive earth pressure by first mobilizing resistance behind the wall and then subjecting the test configuration to shake table excitation. Draw attention tothe potential impact of this often overlooked dynamic passive pressure “backfillinertia effect” on the available force-displacement resistance.6) Develop a method for including the above backfill inertial effect in dynamic bridgesimulations. Demonstrate using FE models.7) Provide new insight into the overall dense sand backfill performance based on thisunprecedented collective set of laboratory and large scale passive and dynamic earthpressure experimental data. 被动位移和力的动态土压力实验英文文献和中文翻译(4):http://www.751com.cn/fanyi/lunwen_24460.html