Behavior of group piles under combined loadings after improvement of liquefiable soil with nanomaterials

1 Introduction

Nanomaterials are a new material level between atoms, molecules, and macroscopic materials with at least one spatial dimension that does not exceed 100 nm. Nanomaterials have a higher specific surface area than their volume, leading to much water encompassing the outer surface. The existence of those materials in soil usually enhances its thixotropic property and increases its shear strength. These materials can be used by reducing the possibility of liquefaction with hydraulic fills or saturated fine sand [1]. Many different types and methods of mitigation were found and suggested for use depending on many factors. The highlights must be focused on ground mitigation, and improvement techniques must be sustainable, keep time, and be cost-effective. The current research focuses on one of the sites containing civilian pile foundations of existing buildings; thus, the target is for mitigation methods that satisfy this condition. During liquefaction, the lateral resistance decreases, whereas the lateral deformation (buckling failure) increases [2]. The liquefaction potential can be reduced by increasing the shear strength and the densification extent. Grouting can be used in constrained sites; traditional grouting materials include cement slurry, sodium silica, and chemical grouts. The chemical grouting method has a high risk of damage to adjacent buildings and pollution of waterways; therefore, with increased industrialization and urbanization and the progress of creative science and technology, alternative materials for mitigating liquefaction risks should be explored [3]. Currently, nanoparticles have been used for stabilizing instead of applying traditional materials; these materials impact physical, chemical, and engineering properties. Hence, using permeation grouting with nanomaterials as a novel liquefaction mitigation method, this method does not cause ground disturbance and can be treated over the entire site. Çelik et al. [4] studied the effects of two types of nanomaterials (nano-aluminum and nanosilica) on the strength properties of poorly graded sand under different percentages of water injected into the sandy soil with a relative density of 70% and a pressure of 2 bars for 7 and 28 days as curing time. It can be concluded that the compressive strength increases with the addition of nanoparticles up to a certain point and stops after a particular point. The liquefaction mitigation methods are dependent on many principles, including densification, reinforcement, and accelerated saturated soil drainage during earthquake motions and other processes. Liquefaction mitigation in the urban area is very complex and has many limitations [5]. Thus, in this study, the focus is to use non-destructive methods in the mitigation process. Permeation grouting with nanomaterials in a horizontal direction is one of the mitigation processes to improve the liquefied soil surrounding the pile foundations. The improvement method occurs by changing the soil skeleton and behavior during liquefaction triggering.

2 Materials used

The sandy soil that was used in this study is classified as poorly graded sand (SP) according to the Unified Soil Classification System (ASTM D2487). The grain size distribution curve of this soil is shown in Figure 1. Two types of nanomaterials (nano-SiO₂ and nano-clay) were used. Tables 1 and 2 show the properties of these materials.

Figure 1

Grain size distribution curve.

The X-ray diffraction analysis of natural and treated sand soil is important for determining changes in the mineralogical composition of the soil as a result of determining whether or not there is a chemical reaction for nanomaterials with sand soil. The curves in Figure 2 included the natural sandy soil, (1/30) nano-clay, and (1/50) nano-SiO₂ obtained from the X-ray test. The intensity of the sand soil mineral peaks decreased slightly because of the used nanomaterials on the sulfate present in the soil, which led to a decrease in the amount of liquefied sand [6]. The X-ray test results given in Figure 2 display the same patterns for the natural and treated sandy soil; this is normal due to the addition of calcium chloride, which is a support reagent that provides calcium, which in turn combines with the carbonates from the hydrolysis to form calcium carbonate. The most prominent clay minerals are calcite, as it was observed that the use in the treatment led to increased calcite precipitation in the sand soil samples. Diffraction analysis of treated soil is important to determine the changes that occur in the mineralogical phase reactions. These reactions depend on the chemical and mineralogical composition of each soil and additive. Quartz X-ray diffraction (XRD) peaks are slightly higher than other mineral peaks because quartz spreads X-rays better than other substances. The peak was identified as quartz at 2-Theta 20.9°. The intensity for sand soil before and after treatment with nanomaterials is shown in Table 3; the intensity increased slightly, reflecting the effect of the nanomaterial’s existence on the sandy soil. These strength changes again emphasized the influence of the added nanomaterials. As compared to untreated soils, nanomaterial-treated soils show many new peaks. The fact that the chemical action of the nanomaterials was approved with the results found from the strength test may explain this result.

Figure 2

X-ray analysis of sand treated with nano-clay and nano-SiO₂.

This involves injecting a grout that permeates the surrounding soil pore space before the grout begins to harden or set. The hardened grout improves the soil by cementing the soil particles together and filling the void space [7]. This primary advantage prevents disturbing the native soil due to associated settlements or distress to an existing foundation or structure. The permeation technique includes grouting nanomaterials (gel–solid) into saturated loosely sandy soil using a permeation time and rate. For this purpose, grouting machine-cell equipment that activated seepage through soil strata of sand was used for grouting the gel nanomaterials. The permeation grouting technique includes grouting nanomaterials (gel–solid) into saturated loosely sandy soil using a permeation time and rate. For this purpose, grouting machine-cell equipment that activated seepage through soil strata of sand was used for grouting the gel nanomaterials. It is locally manufactured from ductile steel to resist high pressure with a grouting cell that connects to the air pressure compressor to permit the grouting materials in the cell tube through a pumping hose. The permeation grouting device was manufactured by AL-Kinani [8]. The gel-like solid was injected into sandy soil strata in a laminar soil box to a specific zone depth of loosely saturated sandy soil with a pressure level of 1.2 psi with a curing time of 48 h and nanomaterial ratios of 1/50 and 1/30 for nano-SiO₂ and nano-clay, respectively, through plastic pipes of 3 mm placed in different locations in the horizontal direction of the loosely saturated layer. The overall observations for soil mitigation based on permeation grouting using nanomaterials were assessed by the pore pressure generation, lateral and vertical displacements, acceleration, and bending moments along the pile depth during the Kobe earthquake. Sensors (e.g., accelerometers, ACC2 and ACC3; pore water pressure ratios, ru1 and ru2; lateral and vertical displacement transducers; and strain gauges in the first and second rows of piles, P1 and P2) were used in this study to record various parameters during shaking table tests. The piles were located at a height of 100 mm above the sandy soil surface. Hence, the total pile length is 500 mm. The total allowable axial load capacity was applied by placing the weight block directly on the pile cap. In order to simulate the lateral loading system on the pile head, an automated pneumatic-vacuum loading system was locally manufactured [9,10]. The total allowable lateral load capacity equals 20% of the total allowable axial load. In this study, eight test cases were considered in the shaking table model. In G-1-K and G-2-K case tests, a pure axial load was applied to the pile, whereas the other test cases included combined (axial and lateral) loadings (G-1-2-K and G-2-2-K) through two types of nanomaterials. Figure 3 shows all details of the setup model.

Figure 3

Setup model.

3 Results and discussion

Figure 4 shows pile cap displacement in both directions, and the first impression to be noted here is that the lateral displacement is improved with the use of nano-clay when subjected to strong motion as well as the lateral loading effect, with 100% lateral loading and 50–100% axial loading. When compared to untreated soil, the lateral displacement of the pile cap will be reduced by about 20% for all test cases: G-1-K, G-1-2-K, G-2-K, and G-2-2-K. On the other hand, increasing cap masses by 50% axial for treated cases G-1-K and G-1-2-K will reduce the vertical 15 to 7 mm and 25 to 10 mm at 100% axial loading for both G-2-K and G-2-2-K, showing a 53% and 60% reduction, respectively, compared to untreated soil.

Figure 4

Vertical and lateral displacements of the pile group in the treated and untreated soil with nano-clay at (a) 50% axial load, (b) 100% axial load, (c) 50% axial load and 100% lateral load, and (d) 100% axial load and 100% lateral load.

The pile response resulting in the middle of treated soil in the shaking table using nano-SiO₂ with 1/50 is shown in Figure 5. With the effect of lateral loading, the liquefied layer around the pile group has shown that the lateral displacements were considered more pronounced for strong shaking motions, showing a surprising increase for the last case G-2-2-K, in which this movement will reduce from 21.59 mm in untreated soil to 14.70 mm in treated with nano-SiO₂. The overall pile lateral displacements in the treated layer were reduced by 30–35% for test cases G-1-K, G-1-2-K, G-2-K, and G-2-2-K when compared to the equivalent measured displacement for the untreated soil. However, due to strong motion, the vertical displacement suffers from soil distribution around the piles in the treated layer and beneath the pile in the untreated layer.

Figure 5

Vertical and lateral displacements of the pile group in the treated and untreated soil with nano-SiO₂ at (a) 50% axial load, (b) 100% axial load, (c) 50% axial load and 100% lateral load, and (d) 100% axial load and 100% lateral load.

Figures 6 and 7 show the moment profiles of the group pile during strong earthquake motion of Kobe with 100% lateral loading and 50–100% axial loading embedded in treated and untreated liquefied soil with nano-clay. The peak bending moments on P1 and P2 varied linearly with depth within the liquefied and dense non-liquefied layers, suggesting that the liquefied layer was so weak that it did not appreciably contribute to the moments of their distribution along with the piles. So, the peak bending values occurred at the upper and lower boundaries of the liquefied and non-liquefied layers during the early part of the shaking, nearly at a depth of 0.75 L below the shaking table surface. As shaking continued, the moment value at the lower boundary of the liquefiable sand layer remained at the interface between the untreated and the treated liquefiable layers. This suggests that the treated soil with nano-clay around the pile with 100% lateral load has lowered the peak bending value on P1 at the interface boundary between the two layers to about 25% for 50% and 100% axial loading due to increases in cyclic strength for nanoparticles and decreases in imposed bending moments along with pile cap. At the same time, these values rounded to about 27% for the second row with the same loading.

Figure 6

Bending moment of the pile group in the treated and untreated soil with nano-clay – P1 at (a) 50% axial load, (b) 50% axial load and 100% lateral load, (c) 100% axial load, and (d) 100% axial load and 100% lateral load.

Figure 7

Bending moment of the pile group in the treated and untreated soil with nano-clay – P2 at (a) 50% axial load, (b) 50% axial load and 100% lateral load, (c) 100% axial load, and (d) 100% axial load and 100% lateral load.

The influence of permeation grouting within loosely liquefied soil that was laid on the upper portion of the shaking table using nano-SiO₂ on the developed bending along pile length is illustrated in Figures 8 and 9; the result indicated that the peak bending was slightly lower at the interface boundary when compared with nano-clay. The peak moments on P1 and P2 also occurred at the upper and lower boundaries between treated and untreated layers of 375 mm. The nano-SiO₂-treated layer enhances the neighboring soil, resulting in the measured moments of P1 within the leading row exhibiting a reduced value ranging from 34 to 38% for increasing axial mass loading from 50 to 100%, respectively, at constant lateral loading of 100%.

Figure 8

Bending moment of the pile group in treated and untreated soil with nano-SiO₂ – P1 at (a) 50% axial load, (b) 50% axial load and 100% lateral load, (c) 100% axial load, and (d) 100% axial load and 100% lateral load.

Figure 9

Bending moment of the pile group in treated and untreated soil with nano-SiO₂ – P2 at (a) 50% axial load, (b) 50% axial load and lateral load, (c) 100% axial load, and (d) 100% axial load and 100% lateral load.

Acceleration time histories ACC2 and ACC3 measured in the soil deposit of the shaking table at a depth of 120 and 220 mm are shown in Figure 10 for treated and untreated layers with nano-clay. The permeation grouting with nano-clay of the shaking table shows that the acceleration records in the near-field of ACC2 and ACC3 for the treated surrounding group of the pile in the upper sand layer have similar behavior to the measured input motions of the untreated layer except that the amplitudes were not symmetrical and exhibited large spikes in positive acceleration amplitude. At the same time, it dropped in the negative direction of shaking for all treated cases with nano-clay than for the untreated ones. This could indicate that the acceleration of the treated layer during shaking may be attenuated due to dampening related to the addition of nano-clay dispersion between sand particles.

Figure 10

Acceleration within different depths of treated and untreated soil with nano-clay: (a and b) ACC2 and ACC3 at 50% axial load and 50% axial load and 100% lateral load; (c and d) ACC2 and ACC3 at 100% axial load and 100% axial load and 100% lateral load.

The acceleration of the treated liquefied sandy layer with nano-SiO₂ is described in Figure 11. From these results, it can be shown that there is a drop in amplitude after several cycles during a strong Kobe earthquake. The magnitude of the accelerations ACC2 and ACC3 was noted to be reduced at the peak of the ground earthquake motion by about 30% for all test cases as compared to the untreated layer, and this scenario was attributed to the effect of nano-SiO₂, which plays an important role in the stiffening and cementation of tiny nanoparticles that tends to less dilative behavior of the nano-SiO₂-treated layer that remained solid after several cycles of shaking. The overall acceleration amplitude of the treated layer with both nano types indicates that grouting with nano-SiO₂ experiences a slight effect over the nano-clay of the liquefied layer due to the stiffness and strength effect of tiny nano-SiO₂ that will solidify the layer after several cycles, and thus, the soil did not liquefy at the first shaking duration.

Figure 11

Acceleration within different depths of treated and untreated soil with nano-SiO₂: (a and b) ACC2 and ACC3 at 50% axial load and 50% axial load and 100% lateral load; (c and d) ACC2 and ACC3 at 100% axial load and 100% axial load and 100% lateral load.

Figure 12 illustrates the values of the pore pressure ratio before and after treating the liquefiable layer with nano-clay. It can be seen that the pore pressure propagation rate for ru1 and ru2 was much lower for the treated layer with nano-clay than in the untreated layer; therefore, the pore pressure ratio for the treated layer gradually accumulated and reduced, reaching a maximum value of 0.65 and 0.5 at peak input motions corresponding to ru1 and ru2, respectively, as compared to that untreated with 0.88 and 0.74. However, the treated layer appears to have lagged behind the earthquake after reaching its peak value long after shaking ceased for about 9–10 s due to the influence of nano-clay in the form of a solid-like gel that does not entirely fluidize. Therefore, the pore pressure generation is delayed. Moreover, the peak values of ru1 and ru2 are lower in the treated layer than in the untreated one. This finding is similar to the results noted by refs [11] and [12].

Figure 12

Pore pressure ratio within treated and untreated soil with nano-clay: (a and b) ru1- and ru2-treated at 50% axial load and 50% axial load and 100% lateral load; (c and d) ru1- and ru2-untreated at 100% axial load and 100% axial load and 100% lateral load.

The mechanisms of soil grouting by nano-SiO₂ to mitigate liquefaction potential are pore fluid solidification, soil grain cementation, and the delay of pore pressure generation. The influence appeared in Figure 13, indicating liquefaction resistance. The pore pressure accumulation process in the treated layer is delayed as presented in recorded values of ru1 and ru2, and the deformation of loose sands subjected to earthquake shaking is much smaller than that of the untreated layer. Also, it must be noted here that all treated and untreated soil will not experience a liquefaction potential since both r _u values will not reach 1 and indicate significant drainage away from the soil beneath the pile cap.

Figure 13

Pore pressure ratio within treated and untreated soil by nano-SiO₂: (a and b) ru1- and ru2-treated at 50% axial load and 50% axial load and 100% lateral load; (c and d) ru1- and ru2-untreated at 100% axial load and 100% axial load and 100% lateral load.

4 Conclusions

1. The reduced value of acceleration at pile cap has resulted from the contribution of the lateral loading effect, so the treated liquefied layer with both nanomaterial types had no effect on the measured acceleration at pile cap.

2. The treated soil did not liquefy and thus transmitted shear stresses through the surface induced by the base shaking. The overall acceleration amplitude of the treated soil layer shows the permeation grouting with nano-SiO₂ experiences a slight effect over nano-clay of the liquefied layer.

3. The inclusion of nano-materials with liquefied soil layer can delay the generation of pore water pressure during the Kobe earthquake by about 9–10 s due to the influence of nano-clay in the form of a solid-like gel that does not entirely fluidize.

4. The maximum bending moments along the pile were concentrated near or around the pile cap and interfaces between the liquefiable and non-liquefiable sand soil layers. The maximum moments were proportional to the permanent lateral ground displacement. The treatment of liquefiable soil with nanomaterials reduced pile bending moments by 30% compared to the moment measured in the untreated model.

5. The lateral pile displacements in the treated soil layer were reduced by 30% for test cases compared to the equivalent measured displacement for the untreated soil, whereas the vertical displacement values of the treated soil displayed a reduction of nearly 50–60%, including 100% lateral loading for both nanomaterials.