Determination of Initial-Shear-Stress Impact on 2 Ramsar-Sand Liquefaction Susceptibility through 3 Monotonic Triaxial Testing 4

: Liquefaction risk assessment is critical for the safety and economics of structures. As the 17 soil strata of Ramsar area in north Iran is mostly composed of poorly graded clean sand and the 18 ground water table is found at shallow depths, it is highly susceptible to liquefaction. In this study, 19 a series of isotropic and anisotropic consolidated undrained triaxial tests are performed on 20 reconstituted specimens of Ramsar sand to identify the liquefaction potential of the area. The 21 specimens are consolidated isotropically to simulate the level ground condition, and anisotropically 22 to simulate the soil condition on a slope and/ or under a structure. The various states of soil behavior 23 are studied by preparing specimens at different initial relative densities and applying different 24 levels of effective stress. The critical state soil mechanics approach for identifying the liquefaction 25 susceptibility is adopted and the observed phenomena are further explained in relation to the micro- 26 mechanical behavior. As only four among the 27 conducted tests did not exhibit liquefactive 27 behavior, Ramsar sand can be qualified as strongly susceptible to liquefaction. Furthermore, it is 28 observed that the pore pressure ratio is a good indication of the liquefaction susceptibility


52
The influence of the initial shear stress on the liquefaction susceptibility has been investigated in 53 recent decades. Kramer and Seed [13] observed that in samples consolidated to principal effective 54 stress ratios (K`) of 1.5, 2.0, and 2.25, the increase in deviator stress under undrained conditions 55 required to initiate liquefaction were approximately 0.6, 0.25, and 0.13 ksc (59, 25, and 13 kPa), 56 respectively. They reported that the resistance to static liquefaction in these samples decreased 57 significantly as the initial shear stress level increased. Later contributions by Harder and Boulanger 58 [14], and Seed and Harder [15] demonstrated that the presence of the initial shear stress ratio 59 improves liquefaction resistance at high relative densities (55-70%), whereas the effect is less 60 pronounced at low relative densities ( approximately 35%). According to recent research [16,17], 61 triaxial tests performed at higher initial shear stress parameter values have a distinct effect for range 62 of α. Alpha values could create both improving and aggravating effects on loose or very dense 63 sands influence on the liquefaction susceptibility.

64
Soil samples under shearing up to large strains tend to reach a state of continuous deformation 65 under constant shear (q) and normal stresses (p'), where such occurrence is known as the ultimate 66 steady-state line (SSL). Soil at SSL exhibit a relationship between the ultimate values of the deviatoric 67 stress and mean effective principal stress. Therefore, soil behavior can be predicted by expressing the 68 state of the effective confining stress and defining the location of this point relative to the steady-state 69 line. Castro and Poulos [4] observed that in addition to the steady-state line position being a unique 70 soil property, the inclination of steady-state lines vary extensively, even for apparently similar soils.

71
The steady-state concept and SSL were further described by other research [1,3,18]. Poulos et al. [4] 72 systematically measured the steady strength through stress-controlled CU triaxial tests. On the other 73 hand, Roscoe et al. [19] studied the yielding of soils and reported that when soil is subjected to shear 74 distortion, it begins to shear at constant volume at a certain critical pressure. Similarly, according to 3 of 18 Schofield and Wroth [20], if a soil specimen is continuously distorted, it reaches steady state, i.e., flow 76 failure. Although the critical state was initially developed on clayey soils, several studies have 77 attempted to adopt this framework to granular materials [21,22]. However, Been et al. [23] stated that 78 such an attempt is challenging due to the difficulties in determining the normal consolidation line for 79 such soils. Coop [24] clarified that this phenomena involves stress formation at the particle contact 80 points, affecting the compression response of granular materials. However, the attempts of Ferreira 81 and Bica [25], Coop [26], Ekinci et al. [12], and Rezaian at al. [27] for adopting this framework to 82 granular soils was successful.

83
To understand the critical state concept of granular materials, the microscale behavior of sandy 84 soils was investigated. Cavarretta et. al. [28] studied the micromechanical behavior of coarse-grained 85 soils utilizing a new technology to compute the particle shape and surface roughness, for measuring 86 the particle contact stiffness and interparticle friction to relate the nature of fundamental particle 87 behavior with the traditional test results (triaxial and oedometer). Although a link was established 88 between the roughness of the particle surface and interparticle friction, the influence of the particle 89 shape was more noticeable. In a similar study by Senetakis et al. [29], repeated interparticle shear 90 testing showed a small decrease in the friction angle, which can be because of asperity damage during 91 initial shearing. Recently, Zhang et. al. [30] investigated sands with a variety of minerology using 92 shape analysis, particle crush tests, and one-dimensional compression tests. Authors reported that 93 particle mineralogy could be a major factor affecting the strength or compressibility, rather than the 94 particle shape. Moreover, Zhao et. al. [31] investigated the effect of the initial density of specimens in 95 one-dimensional compression to evaluate particle breakage. They stated that specimens prepared 96 with high relative density had lower probability of failure and failure modes that were less extensive 97 compared to the low-density specimens. Moreover, it was reported that the effect of the initial density 98 on the probability of particle survival reduced after significant breakage. Additionally, loose 99 specimens exhibited higher compression due to particle failure leading to more fine generation by 100 the further crushing of the existing fragments.

101
In support of the anisotropical soil condition on a slope and/ or under a structure, McDowell 102 and Bolton [32] proved that compared to isotropic or k0 conditions, shearing was more effective at 103 breaking particles. Furthermore, Coop et. al. [33] investigated particle breakage by performing ring 104 shear tests and observed that particle breakage continued up to very large strains along with 105 volumetric compression, which was observed even for tests at moderate confining stresses.

106
Accordingly, Chandler [34] stated that the observed critical state at the strain levels reached by triaxial 107 equipment was due to the counteracting dilative strains because of particle rearrangement and 108 compressive strains because of particle breakage.

116
In this study, the effect of the initial shear stress ratio on the potential liquefaction susceptibility 117 is investigated through monotonic testing. This study is the first to evaluate the liquefaction 118 susceptibility by relating the pore pressure ratio (ru) with the initial shear stress ratio (α) and explain 119 the liquefaction phenomena in relation to the particle breakage mechanism. Furthermore, the 120 adopted anisotropic testing mimics realistic scenarios such as the simulation of the soil condition on  The relative density Dr is a critical parameter that can control the stress-strain behavior or change 131 the liquefaction susceptibility. In this study, three values of the relative density (2%, 30%, and 45%) 132 were considered to include very loose, loose, and medium dense sand, respectively. Ramsar sand is 133 poorly graded clean sand which is extensively found in the southern coast of the Caspian Sea and

142
The grain size distribution of the specimens after triaxial testing were investigated and plotted 143 against sand specimens obtained at 2-m depth from a borrow pit. Figure 2 shows that sand particles 144 around 0.6-0.8 mm are further broken because there is a reduction in this particle percentage,

145
whereas adversely, there is an increase in the particle percentage in the 0.1-0.2 mm particle size 146 range. This observation reveals that after shearing, Ramsar sand particles are brittle.

147
In order to further investigate this observation, sand particles before and after the test were 148 examined using an optical microscope. The grains of the sand in Figure

156
Furthermore, in order to assess the influence of the observed phenomenon during undrained 157 shearing, two tests under the same confining pressure and relative density (σ3c = 350 kPa and Dr0 = 158 30 %) were conducted on broken and unbroken sand. In these two tests, the ultimate deviatoric stress 159 (qu) for broken sand was higher (283.85 kPa) than that for the unbroken one (255.9 kPa), indicating 160 that the former has more resistance to contractive behavior. Similar to these findings, Cavarretta et.
[28] had reported that compared to the as supplied particles (perfectly rounded glass ballotini),

162
crushed particles showed higher deviatoric stresses and more dilation.
Moist tamping, which is the most common and valid method, was used to prepare the 165 specimens. Sand specimens were oven-dried, mixed with 5% distilled water, and divided into five    The SSL for the 27 CU tests of Ramsar sand are displayed in Figure 6.

205
[12] reveals that the densities of the specimens in this study falling into Group which is located at a

238
The excess pore water pressure (PWP) buildup of three specimens with a relative density of 45%,

239
confining pressure of 150 kPa, and different initial shear stress (α=0, 0.3, and 0.5) are shown in Figure   240 8. All the samples in these three tests exhibit dilative behaviors because of the high relative densities,

241
whereas their PWP intensities differ depending on the initial shear stress ratio. In addition, on 242 increasing the initial shear stress, the PWP peak declines and subsequently, the intensity of dilation 243 increases. Due to the reduction in the excess pore pressure, the particle contact friction increases and 244 causes resistance to liquefaction.

259
The observations in Figures 8 and 9 regarding the effect of confinement on the excess pore 260 pressure, on varying the initial shear stress ratios, can be a criterion for liquefaction susceptibility.

261
From equation 2, it can be observed that the criterion for liquefaction susceptibility pore water 262 pressure ratio (ru) is the variation of the pore water pressure at failure to the initial effective confining 263 pressure.

265
The susceptibility ranges from 0-1, when ru approximates unity, and the liquefaction potential 266 increases. Jafarian et al. [5] had also stated that it is common to have lower ru within the liquefaction 267 phenomenon, termed as susceptible to liquefaction. Figure

276
In Figure 10, the scattered pattern of points shows that ru is related to the relative density (Dr)

293
The friction angle at the failure point and ru have a negative correlation, as depicted in Figure   294 12 3.4. Pore water pressure ratio versus the initial shear stress ratio 302 Figure 13 displays the relationship between ru and α in three charts, whose effective confining 303 pressures are different. It is obvious from the charts that on increasing the relative density, the 304 liquefaction susceptibility reduces regardless of whether α and σ3c increase or decrease. This indicates 305 that the relative density is much more effective than α and σ3c. Figure 13(a) shows that the initial 306 shear stress is not effective in loose sand (2%), whereas it influences semi-dense and dense specimens.

307
In Figure 13(b), this pattern slightly changes because the relative density after consolidation increases, 308 which is strongly related to the effective confining pressure for loose sand. For example, as shown in 309

338
However, at high effective confining pressure, ru values barely reduces as a result of increasing the 339 initial shear stress ratio.

340
The pore water pressure ratio ru is a key parameter for evaluating sand behavior.