Influence of clay content on repeated liquefaction resistance of sand using element test

Soil liquefaction is a major concern because of its potential to cause building collapse, roadbed settlement, and sand eruptions. Field investigations have shown that clayey sand, when liquefied during mainshock, may undergo multiple liquefaction events during the subsequent aftershocks. Although previous research has primarily focused on the initial liquefaction event in clayey sand, research on the mechanism responsible for repeated liquefaction is limited. Consequently, a series of cyclic triaxial experiments are conducted on sand with varying clay contents, and these samples are subjected to various earthquake sequences. The primary objective of this study is to evaluate the cyclic behavior of sandy soil with clay content and analyze its repeated liquefaction resistance under different load sequences. A comprehensive assessment is conducted to understand the evolution of the repeated liquefaction resistance. The test results indicate that clayey sand exhibits the lowest liquefaction resistance during the second liquefaction event. Repeated liquefaction is dominated by stress-induced anisotropy at earlier events and relative density at the final events.


Introduction
In recent years, the frequency of earthquakes has increased worldwide.One major concern associated with earthquakes is soil liquefaction, which can cause building collapse, pile foundation shifts, and even slope sliding [1][2].A seismic survey report revealed that sandy sediments that liquefied during the main earthquake may undergo liquefaction again during subsequent aftershocks, a phenomenon known as repeated liquefaction [3].Remarkably, even when the aftershock was smaller in magnitude than the main shock, the damage induced by repeated liquefaction may be greater than that of the initial liquefaction.For instance, in the eastern suburbs of Christchurch in 2011, the phenomenon of two adjacent liquefactions and sand blasting from the same building crack implied that the volume of ejections resulting from the aftershock exceeded that induced by the mainshock [4].Likewise, during the Great East Japan Earthquake in 2011, at least 90 deposits experienced liquefaction [5].Therefore, the responses and mechanisms of repeated liquefaction have attracted increasing attention.
To study the mechanism and mechanical behavior of repeated liquefaction, Finn et al. [6] conducted triaxial shear tests and observed that the resistance of the soil experiencing the first liquefaction significantly decreased during the second loading.Ha et al. [7] and Ye et al. [8] conducted shaking table tests and found that the liquefaction resistance decreased in the second liquefaction event, but gradually increased in subsequent liquefaction events.In addition, Ni et al. [9] demonstrated that a large residual axial strain significantly affects the liquefaction resistance.Current studies on repeated liquefaction have mainly focused on main and secondary liquefaction events, with limited research on further liquefaction cases.Thus, it remains unknown how and when liquefaction resistance improves during continuous 1334 (2024) 012048 IOP Publishing doi:10.1088/1755-1315/1334/1/012048 2 liquefaction events.In addition, field investigations have shown that repeated liquefaction commonly occurs in sand with clay content [10][11].The clay content affects the interconnection of the soil particles and fundamental properties of the soil [12].Consequently, the cyclic response and resistance of clean and clayey sands differed significantly.Currently, there are limited publications on the repeated liquefaction behavior of clayey sand.
This study intends to investigate the influence of clay content on repeated liquefaction resistance.First, a series of cyclic triaxial tests was conducted on clayey sand with varying clay content to examine the initial liquefaction response.The samples were then tested with different seismic sequences to investigate the evolution of repeated liquefaction resistance.Finally, the repeated liquefaction response of the clayey sand was analyzed carefully.

Material and testing equipment
In this study, clayey sand was prepared by mixing Fujian sand and Shanghai clay (silty marine sedimentary clay).Fujian sand, characterized by angular and subangular particles, is commonly used in sand-liquefaction experiments.Shanghai clay was obtained from an excavation site in Shanghai.Various amounts of Shanghai clay particles were mixed to attain target clay contents (CC) of 0%, 5%, and 10%.The physical properties of Fujian sand, Shanghai clay, and their binary mixtures are listed in tables 1 and 2, respectively.Figure 1 presents the grain size distribution curves for the Fujian sand and Shanghai clay.
Table 1 According to Thevanayagam et al. [13][14] and Rahman et al. [12], as the CC increases, the particle contact state in sandy soils varies.These binary mixtures exhibited the following e max and e min values: for CC=0%, 5%, and 10%, e max and e min were 1.007 and 0.626, 1.028 and 0.560, and 1.084 and 0.494, respectively.
An electronically controlled pneumatic dynamic triaxial loading apparatus was used to conduct repeated liquefaction tests on clayey sand.This dynamic triaxial system allows cyclic loading frequencies ranging from 0.01 Hz to 10 Hz.The system includes an axial stress sensor, axial

Specimen preparation and testing procedure
The Fujian sand and Shanghai clay were crushed, sieved, and dried.The dry soil mass required for each layer was calculated based on the relative density (D r = 55%).Dried sand, clay, and 5% mass fraction of de-aired water were weighed and mixed until no noticeable soil aggregates remained.Standard specimens with a diameter of 50 mm and a height of 100 mm were prepared using the moist tamping method.The soil was filled in a copper mold (figure 2a-b) in five layers, and each layer was uniformly compacted to the target height using a flat-bottomed rammer.Finally, the copper mold was removed (figure 2c).To saturate the specimens, CO 2 and de-aired water were successively injected into the samples and a back pressure of 200 kPa was employed to dissolve the remaining CO 2 .The values of Skempton B were greater than 0.95 for all the specimens in this study.
After saturation, isotropically drained consolidation testing was performed while maintaining an effective stress p o at 100 kPa.Next, a cyclic stress amplitude (q cyc ) of 30 kPa was applied to the top of the specimen using the cycle number (N f ) corresponding to a double axial strain of 5% as the criterion for liquefaction.Subsequently, for the specimens that experienced liquefaction, loading was terminated once the deviatoric stress reached zero (q = 0, p = 0).Subsequently, the liquefied sand was reconsolidated at 100 kPa.As demonstrated by Ni et al. [15], the residual axial train (ε a R ) has great influence on the repeated liquefaction response.Consequently, the values of ε a R are predominantly observed in an expanded state on the extension side, where the specimens initially experienced zero deviatory stress.

Initial liquefaction of sand with different clay contents
Three sets of cyclic triaxial tests were conducted to demonstrate the influence of clay content on the liquefaction behavior.Figure 3 illustrates the cyclic behavior of Fujian sand with different clay contents during the initial liquefaction process.As shown in figure 3(a), the cyclic behavior of clean sand (CC=0) was observed in three stages.In the initial stage, during early loading, the mean effective stress (p) decreased with an increasing number of cycles, accompanied by a gradual accumulation of axial strain and excess pore water pressure (EPWP).Subsequently, the rate of p declines and EPWP accumulation slows, leading to a prolonged development phase.Finally, by the 45 th cycle, the stress path exhibits a "butterfly" shape, with repeated shear dilatancy and contraction on the compression and extension sides, accompanied by a rapid development of axial strain and EPWP.
Compared with clean sand, an increase in the clay content gradually increased the accumulation of axial strain on the extension side.The presence of clay affected the microstructure of the clayey sand, resulting in distinct liquefaction strength characteristics.Figure 3 shows that the clayey sand exhibited the lowest resistance to liquefaction when the clay content reached 10%.In contrast to the cyclic activity exhibited by clean sand and 5% clayey sand, this sand displayed a flow liquefaction mode.This can be attributed to the relatively low clay content, with sand particles dominating the behavior, whereas clay particles were mainly found in the pores between the sand grains.As the cyclic loading progressed, the clay particles acted as lubricants between the sand grains, reducing the liquefaction resistance of the clayey sand.

Repeated liquefaction of sand with different clay contents
This section evaluates the repeated liquefaction behavior of clayey sand, involving repeated consolidation and cyclic loading after the initial liquefaction.The anisotropy generated during liquefaction can be observed as the difference between the compressive and extensive strains, with a greater difference signifying a more pronounced anisotropy [17], proposed as the relative residual axial strain ( ).As shown in figure 4, the stress paths for sand with a large negative relative residual axial strain are similar in the first cycle.During this cycle, the mean effective stress decreases sharply and moves to the cyclic mobility stage.Furthermore, EPWP developed rapidly during the first cycle.It is notable that the repeated liquefaction resistance of sand specimens with large negative relative residual strains mainly depends on the accumulation of axial strain during continuous loading events, as evidenced by the stress-strain relationships in figure 4. The accumulation of the axial strain gradually slowed as the number of liquefaction events increased.Moreover, the stress paths were significantly influenced by the relative residual axial strain from the previous liquefaction events.Figure 5 illustrates the correlation between the normalization of the number of cycles required for liquefaction, the occurrence of the first liquefaction case, and the number of liquefaction events.Overall, the second liquefaction event exhibited the lowest liquefaction resistance for sand regardless of the presence of clay.The normalized liquefaction resistances remained relatively constant from the second to sixth liquefaction events.However, sand with higher clay content regained its strength during an earlier event.Compared with the liquefaction resistance at the first liquefaction event, clayey sand with clay contents of 0%, 5%, and 10% experienced liquefaction events of 9, 10, and 8%, respectively, to regain liquefaction resistance.During the reconsolidation process, the EPWP dissipated, and D r increased, traditionally causing the connections between the soil particles to become tighter.As shown in figure 6, the relative density increased gradually with an increase in the number of liquefaction cases.Comparing figure 6 with figure 5, it can be observed that the resistance to the second liquefaction is mainly influenced by the residual axial strain (that is, stress-induced anisotropy [9]), with little impact from D r .Additionally, after the sixth liquefaction event, when the relative density exceeds 75%, the resistance to liquefaction gradually recovers despite having relatively low Δ.The rate of axial strain accumulation decreased gradually, narrowing the difference between the compression and extension strains.Specifically, liquefaction resistance was primarily influenced by Δ in the first six liquefaction events.After these six events, the resistance was predominantly affected by D r while the impact of Δ gradually diminished.As a result, the resistance to liquefaction gradually improved and strengthened beginning from the sixth liquefaction event, leading to increased stability.

Conclusions
This study aimed to investigate the cyclic behavior of clayey sand during repeated liquefaction events.Three clay contents, 0, 5, and 10%, and several continuous cyclic loading cases were considered in the cyclic triaxial tests.The main conclusions are as follows: (a) The stress paths are largely influenced by the relative residual axial strain (Δ) from previous liquefaction events.The repeated liquefaction resistance for sand with a large negative Δ mainly depends on the accumulation of the axial strain during continuous loading events.(b) In the continuous loading liquefaction test cases, the second liquefaction event exhibited the lowest liquefaction resistance for sand, regardless of the presence of clay.For sand with large stressinduced anisotropy (or relative residual strain history, Δ), the normalized liquefaction resistances remained relatively constant during the second to sixth liquefaction event.Subsequently, the sand regained strength compared to the first liquefaction resistance.(c) For sand with a large negative Δ, the repeated liquefaction resistance is mainly influenced by the residual axial strain in the initial sixth liquefaction events, with little impact from the relative density.Subsequently, the repeated liquefaction resistance was predominantly affected by the relative density while the impact of Δ gradually diminished.
pore water pressure sensor, volume sensor, and other relevant components.

Figure 3 .
Figure 3. Cyclic response of sand with different clay content during the first liquefaction event.

Figure 5 .
Figure 5. Relationship between liquefaction events and liquefaction resistance.

Figure 6 .
Figure 6.Relative densities of soil after each liquefaction event.

.
Physical properties of Fujian sand.

Table 2 .
Physical properties of Shanghai clay.