(Figure 1.11).Cole and Rollins (2006) performed cyclic load tests on a 1.1 meter tall, 5.2 meterlong, and 3.1 meter wide pile cap, supported by 12 piles, using four different backfillsoils. Tests were conducted with and without soil backfill in front of the pile cap in order to approximate the contribution of lateral resistance which was provided by the passive pressure (Rollins and Cole 2006). From those tests, Cole and Rollins (2006) concluded that the Log Spiral theory with 3D correction (Brinch-Hansen 1966) provided the best estimates of the peak passive pressure and that the hyperbolic model proposed by Duncan and Mokwa (2001) provided the best agreement with the load-deflection behavior for monotonic loadings. Bozorgzadeh (2007) performed 5 tests on 1.5, 1.7 and 2.3 meter tall and 4.7 and 5.5 meter long bridge abutment walls with silty sand (SM) and clayey sand (SC) backfills. In some of those tests, Bozorgzadeh (2007) constructed the backfill on a cut slope, similar to the construction of a bridge abutment in practice, and found that this method can introduce a weak surface, reducing the peak passive resistance. Bozorgzadeh (2007) also observed significant post-peak strain softening in tests where the wall moved up with the adjacent backfill soil, and proposed a force-displacement model to account for this effect. Similar to the above experimental studies, the Log Spiral method providedgood estimates of the peak passive resistance. In another bridge abutment investigation, Lemnitzer et al. (2009) performed a test on a 1.7 meter tall and 4.6 meter long wall, with well graded silty sand (SW-SM) backfill, by applying lateral and diagonal (downward) loads to provide a purely horizontal wall translation. In that configuration, δmob = 14 degrees at the instant of the peak passive resistance was determined from the vertical component of the diagonal actuator force (Stewart et al. 2007, Lemnitzer et al. 2009). Friction on the bottom of the wall was also measured by performing a test without backfill soil, in order to remove this portion from the estimated abutment passive resistance (Lemnitzer et al. 2009, Stewart et al. 2007).
Lemnitzer et al. (2009) concluded that the peak resistance was well estimated by the Log Spiral method, and the shape of the hyperbolic curves of Eqs. (1.5) and (1.6) provided a good match with the recorded load-displacement behavior.1.1.4 Numerical SimulationsMartin and Yan (1995) conducted numerical studies using the finite differencecode, FLAC, for passive pressure behind bridge abutments. In that study, a soil modelusing the Mohr-Coulomb failure criterion, and an elastic-perfectly plastic constitutive relationship were employed. Martin and Yan (1995) concluded that the FLAC modelsprovided reasonable results in terms of peak passive resistance when compared with the theoretical predictions. More recently, a Hardening Soil (HS) model in the FE program Plaxis (2004), has been demonstrated to provide an improved representation of the passive pressure forcedisplacement relationship up to the peak resistance, compared with experimental results (Shamsabadi and Nordal 2006, Bozorgzadeh 2007). Similar to the Martin and Yan (1995), the HS model employs the Mohr-Coulomb failure criterion, but it improves on the constitutive end by using a hyperbolic stress-strain relationship (Plaxis 2004).1.1.5 Passive Pressure SummaryClassical passive earth pressure theories allow for estimation of the peakresistance as a wall moves toward the backfill, but those predictions do not include arepresentation of the load-displacement relationship. Renewed interest in thedevelopment of passive pressure with displacement has motivated recent excellentexperimental studies. Yet variability of backfill soil shear strength, loadingconfigurations, and wall heights in the field continue to motivate further experimentation.1.2 Dynamic Earth Pressure and the Retaining Wall-Backfill Response toEarthquake ExcitationWhen subjected to earthquake excitation, the dynamic backfill soil response, theinertia and motion of the wall, and the interaction between them can be extremelydifficult to predict. Inertial forces and motions induced by earthquakes may increase the demand on structures by imposing larger forces compared to the static active or at-rest earth pressure conditions (Kramer 1996). Stability of a retaining wall may also be reduced due to the decrease in resisting passive earth pressure (Kramer 1996).1.2.1 Dynamic Earth Pressure TheoriesAs an approach to aid in the design and analysis of retaining structures for seismicloads, two well-known theoretical methods have been in use for predicting dynamic earth pressure. The Mononobe-Okabe (Okabe 1926, Mononobe and Matsuo 1929) equations are the most widely used for so-called “yielding” walls, which are walls that can displace adequately to achieve minimum active or maximum passive conditions. An elastic solution proposed by Wood (1973), considers “non-yielding” walls, such as basement walls, which are assumed to be unable to move significantly to achieve the active or passive state. These two methods for predicting dynamic earth pressure are discussed in detail further below. Additional analyses have been conducted on dynamic earth pressure, including a pseudo-dynamic approach (Steedman and Zeng, 1990), which can account for phase difference and amplification effects (Kramer 1996). Richard and Elms (1979), and Whitman and Liao (1985) also proposed methods which focus on the retaining wall1.2.2 Observed Field Performance of Retaining Walls during EarthquakesWhile damage to retaining walls has been observed after some earthquakes, it hasoften involved a weak (for instance liquefiable) underlying layer (Gazetas et al. 2004,Shirato et al. 2006, Al Atik and Sitar 2008). In the absence of such a weak layer, many retaining structures have performed well, even in cases where the seismic load was not explicitly a design consideration (Seed and Whitman 1970, Lew et al. 1995, Gazetas et al. 2004, Al Atik and Sitar 2008). However in some cases, retaining walls supporting non-saturated backfills have failed or been damaged. For instance after the 1995 Kobe earthquake, Gazetas et al. (2004) reported that several masonry and unreinforced gravity walls were heavily damaged, while reinforced concrete walls experienced relatively little harm. After the 1971 San Fernando earthquake, Clough and Fragaszy (1977) found that U-shaped channel floodway structures designed only for static Rankine (1857) active pressures, performed well with peak excitation up to about 0.5 g, but sustained damage for larger acceleration levels. After the 1999 Chi-Chi earthquake, Fang et al. (2003) also reported on the failure of three gravity walls.1.2.3 Experimental StudiesOver the years, a wide range of shake table experiments have been performed inorder to measure dynamic earth pressure and investigate the retaining wall response 被动位移和力的动态土压力实验英文文献和中文翻译(3):http://www.youerw.com/fanyi/lunwen_24460.html