1. Introduction
In Central Asia, particularly Kazakhstan, almost a third of all areas are formed on saline silty-clayey soils (
Figure 1). Salts in the soils of arid regions are compiled due to the following factors: hot air, penetrating into cracks and pores; an aggressive water–salt soil medium, causing intensive evaporation of the solutions contained in them. Salts that precipitate from the solution in the form of crystalline inclusions fill the pore space of clay soils. The concentration of salts in the solid phase has been observed in the aeration zone to a depth of 3 to 5 m [
1].
The construction of buildings and structures on saline silty-clayey soils in accordance with the regulations NTP RK 07-01.1-2011 “Design of buildings on salt soils” [
2] is associated with an increase in material and labor costs by 5–25% when compared with construction on nonsaline soils. Despite the enormous additional costs, the experience of their operation indicates that many facilities undergo uneven collapsing settlements, exceeding the designed ones by 1.5–3 times and more (NTP RK 07-01.1-2011, 2011) [
2]. According to the All-Russian Research Institute of Hydraulic Engineering named after B.Ye. Vedenev, in 350 out of 960 surveyed irrigation and drainage and hydraulic structures on the saline clayey soils in the republics of Central Asia and Kazakhstan, excess settlements were observed. Soil stabilization methods used in Kazakhstan include soil treatment with cement mortars and silicate solutions, thermal and electrochemical methods, and strengthening with clay soils for noncohesive soils [
3]. However, corrosion and, consequently, a gradual decrease in the bearing capacity of concrete structures in a chemically aggressive soil environment also have great impacts on subsequent operation.
Corrosion-resistant materials for the surface protection of concrete structures, according to review studies of Pan et al., 2017 [
4] and Brito and Kurda, 2021 [
5], in terms of the mechanism of action, can be divided into four groups: surface coating, surface treatments, pore-blocking surface treatment, and hydrophobic impregnation. In terms of corrosion-resistant materials, they are subdivided to organic and inorganic based. The organic protection of a concrete pile by oil-bituminous rocks or natural bitumen is an efficient but labor-intensive and a short-service-life method [
6,
7]. A number of scientists have presented surface coatings with inorganic materials: acrylic [
8], polyurethane [
9], polymer nanocomposite as a polymer-clay [
10], polymer-silica [
11], and epoxy resins [
12].
Hydrophobic impregnation, preventing the penetration of aggressive liquid through the pores of concrete, was implemented by incorporating nanoparticles [
13,
14], micro silica [
15], or stearic acid emulsion [
16]. Increased hydrophobic coatings were also achieved by including ammonium polyphosphate [
17], calcium carbonate nanoparticles [
18], candle soot [
19,
20], cyanoacrylates [
21], and rise husk ash [
22].
The strategies of pore-blocking concrete surface treatment with fluosilicate [
23], sodium silicate [
24], and calcium silicate-based solutions [
25] are some of the types of surface treatment of reinforced concrete structures. However, it should be noted that the high friction of soil particles increases the risk of surface abrasion, and the inability to control the integrity of the protective coating during the operation of the foundations complicate the technology of their application in deep foundations.
As for the ground improvement, the samples treated with silicate cement [
26], geopolymers [
27], and fly ash and slag [
28] have shown reliable results in terms of collapsible deformations under loading in wetting conditions and soil stabilization in general. Thus, based on the experience of previous research, including personal contributions of the authors in this area [
29], it was required to develop a progressive corrosion protection method that meets the design requirements of an aggressive saline soil environment satisfying SP RK 2.01-101-2013 [
30] and, at the same time, increasing the bearing capacity and durability of bored pile foundations:
- -
Formation of a protective corrosion-resistant “shell” around the pile;
- -
Formation of a compacted or ground-improved soil layer around the pile by increasing its strength and deformation properties, structural and suffusion resistance, and water resistance and waterproofness;
- -
Increasing the strength and corrosion resistance of the concrete of the structure by using modifiers, etc.
2. Materials and Methods
2.1. Design of the Protective Case
The formation of a suffusion-resistant waterproof layer of soil with increased strength under the tip and along the lateral surface of the pile when installing a bored pile was achieved by silicatization of the borehole during the installation (
Figure 2 and
Figure 3). For this, a borehole was pre-drilled in saline silty-clayey soils to the required depth from which excess soil was extracted. The inner walls of the borehole were impregnated with a low-concentration sodium silicate solution. The chemical interaction of sodium silicate with carbonates and gypsum contained in saline silty-clayey soils leads to the formation of a viscous silicic acid hydrogel film, which cures on particles and soil aggregates and tightly clogs capillaries and pores. The formation of a silicic acid gel was confirmed by fluoroscopy and microscopy.
Saline silty-clayey soils contain up to 15–30% or more salt inclusions in the form of carbonic acid and sulfuric acid calcium, magnesium, etc. The chemical process of treating the clayey soils around the pile is based on the good penetration of silicate solution into silty clay and on the rapid release of the cementing soil silicic acid gel films due to the reaction of a sodium silicate solution with soil salts [
31]:
The coagulation of a silicate solution in contact with salt waste was confirmed by laboratory testing on the interaction with pure salts (CaCO3, CaSO4·2H2O, etc.) and led to the following conclusions:
- -
The rate of coagulation of silicate solutions and the formation of SiO₂ gel films on salts depend on three factors: the concentration of the sodium silicate solution, and the amount and quality of salts present in the salt waste. Film formation during the interaction of a sodium silicate solution with calcium sulfate slows down with an increase in the concentration of the silicate solution, which is explained by a decrease in the solubility of the reacting salt with the increase in the concentration of silicate. The film thicknesses when using sodium silicate of 10%, 20%, and 30% concentrations were, respectively, 40, 20, and 10 microns.
- -
With regard to the coagulation of silicate solutions in the soil, the sulfuric acid salts of Ca+2; and Mg+2 are of the greatest importance as the most soluble compounds, in comparison with carbon dioxide Ca, which is almost inert.
In addition to the formation of a gel of silicic acid and lime hydrate, exchange reactions occurred in the absorbing complex of the salt waste, and as a result, secondary products formed in an alkaline medium. Therefore, the strength of the fixed salt waste largely depends on the value of the absorption capacity of the salt waste in the natural state, which is expressed in mmol per 100 g of dry salt waste, in an alkaline solution.
In the course of interaction within the porous media, a silicic acid gel film formed on the soil particles and capillary walls. Pores and capillaries in the nearby soil became clogged and impermeable. As the solution was injected, the thickness of the protective casing increased and the strength of the soil rose. The transition of the solution into the gel proceeded during the first 3 days, and the hardening of the gel and, accordingly, the soil mass around the pile proceeded during the next 15 days and ended after 28 days (R = 1.1–1.8 MPa).
Then, the borehole was filled with corrosion-resistant concrete, which was achieved through modified additives that also increased the density of the concrete. Further curing of the casted concrete pile took place in a structurally and suffusion-resistant shell with increased strength, water resistance, and waterproofness. The total diameter of a bored pile casted in a protective load-bearing shell including a silicate-fixed soil layer around the pile was about 600–700 mm.
2.2. Field Tests
Field tests were carried out on saline silty-clayey soils with a content of readily soluble salts up to 3% and carbonates up to 27% for the object: “Retaining wall on Mount Koktobe in Almaty” (
Figure 4). According to the data of engineering-geological surveys carried out at the site with soil sampling to a depth of 20 m, the geological structure of the site’s soils is represented by silty clays interlayered with clayey soils. The content of carbonates in loams is up to 27%, and easily and medium-soluble salts up to 3%. The topsoil layer is 0.3 m. The degree of aggressiveness of soils in relation to the impact on reinforced concrete structures varies from average on Portland cements to strong on sulfate-resistant brands of cement. The physical and mechanical properties of saline silty-clayey soils of the base of the test site are shown in
Table 1.
To study the effect of the desalinization processes on the bearing capacity of the piles, it was decided to install a friction pile with prolonged wetting, as well as a control pile without wetting. To determine the real behavior of the “pile-shell-soil” system, static load tests were carried out with piles in a protective and load-bearing shell for vertical indentation load: in the natural state and during leaching under flooding conditions. Six testing zones were observed during the installation of bored friction piles with a diameter of 400 mm and a length of 4 m. The allocation of the tested piles in each of the sections is presented in
Figure 5. For a comparative analysis of the bearing capacity, near each experimental pile in the soil shell treated with silica solution (hereinafter referred to as “protective shell”), ordinary bored piles without a protective shell (hereinafter referred to as “ordinary pile”) were casted.
Thus, the tests were carried out for:
- -
Six bored piles with a protective shell, without soil wetting;
- -
Six ordinary bored piles, without soil wetting;
- -
Six bored piles with a protective shell, after wetting the soil;
- -
Six ordinary bored piles, after wetting the soil.
Wetting of the soil mass around the pile continued for six months by pumping water into special drainage wells, 0.2 m in diameter, 3.5 m long, located at a distance of 1 m from the edge of the experimental pile (
Figure 5). At the end of the wetting, significant desalinization was achieved near the pile soil massif as a result of suffusion processes and the partial migration of clay particles into the holes of the drainage tubes.
2.3. Site Testing Procedure
Static load tests of bored piles were carried out according to GOST 5686-2012 “Field test methods for piles” [
32] to determine the settlement and bearing capacity of ordinary piles and piles with a protective shell. For testing the piles with static vertical-indentation loads, an anchor-stop stand was used, consisting of systems of main and auxiliary beams, four anchor piles, and anchor ties (clamps).
During the static piles load testing, the following equipment was used:
- -
Hydraulic jack of the SMZh-158A brand with a lifting capacity of 2000 kN;
- -
Manual pumping station MNSR-400 with pressure gauge MTP-160, with a capacity of 785 MPa;
- -
Two deflection indicators 6PAO (for determining the settlement of the pile).
Reactive forces from the step loading of the pile by the jack were perceived by an anchor-stop stand, consisting of four anchor bored piles resisting the tension forces and a stop structure of metal beams (
Figure 6).
Static load tests were carried out by maintained load with the following sequence:
- -
The piles were maintained for 7–10 days from the moment of their installation (until the concrete of the pile reached more than 80% of the design strength);
- -
Loading of the pile was implemented by a vertical-compressing multiple-increasing load and held constant until the rate of settlement fell below 0.1 mm per hour;
- -
Pile settlement was recorded using electrical displacement transducers according to the standard;
- -
Unloading of the tested piles was carried out in steps equal to double loading steps;
- -
Readings of the elastic deformation of the soil and concrete pile during unloading were registered at each stage every 15 min.
3. Results
From the SLT results, the curves between the settlement ratio and load (s/P), where s is the settlement and P is the applied load for all types of piles in the six testing zones, are presented in
Figure 7,
Figure 8,
Figure 9 and
Figure 10. Thus, the tested piles in a protective shell, which are casted in a preliminary treated with sodium silicate solution soil, are presented in
Figure 7. Here, pile #1 is a pile tested in zone 1, pile #2 in zone 2, and so on.
The values specified in the Standard p.p. 5.3–5.5 SP RK 5.01–103–2013 “Pile foundations” (2013) were taken as the criterion for the maximum permissible settlement. By the particular value of the limiting resistance of the pile to the compression load, Fu was understood as the load under which the reaction of the test pile takes a settlement equal to S, determined by the formula:
where
Su,mt is the limiting value of the average settlement of the foundation of the designed structures, where industrial and civil one-storey and multi-storey buildings with a full frame is taken as equal to 8 cm (for reinforced concrete structures) according to the instructions of SP RK 5.01-103-2013 Pile foundations [
33].
ζ is the coefficient of transition from the limiting value of the average settlement of the foundation of structures
Su,mt to the settlement of the pile, obtained during static load tests with conditional stabilization (attenuation) of the settlement, taken as equal to 0.2 according to the instructions of SP RK 5.01-103-2013 “Pile foundations” [
33].
The maximum permissible settlement is accepted to be 16 mm. The particular values of the experimental bearing capacities of the piles corresponding to the maximum permissible settlement are presented in
Table 2.
The results of the statistical check satisfy the condition |Fn − Fi| > vS; therefore, all the obtained results of the carrying capacities can be considered valid and suitable for further analysis.
Analysis of the influence of suffusion processes on the strength characteristics of soils was carried out for both types of pile loadings (
Table 3 and
Table 4). A comparison was made between the bearing capacity of piles with the protective shell and the bearing capacity of ordinary piles after prolonged wetting. The functional linear dependence
to
is expressed by the
k coefficient.
where
is the bearing capacity of piles according to the SLT without soil wetting, kN; and
is the bearing capacity of piles after prolonged wetting of soils, kN.
The relationship between the bearing capacity of piles without and after wetting for piles with the protective shell and for ordinary pile is presented in
Figure 10 and
Figure 11. Regression analysis of the functional dependence was performed using the least squares method. The bearing capacities were obtained from the coordinates of points on the plane, after which the line of best fit was drawn characterizing the proportional relationship, which deviated least of all from these points.
To find the linear equation of the best-fit line using the least squares method, the coordinates of the points
and
were drawn until they intersected with the desired straight line (
Figure 10 and
Figure 11). The values of these ordinates are equal to (
. The ordinate distance from the point
to the straight line is (
. Thus, the regression is expressed by the following equation:
It is assumed that the best-fit line would be the smallest value of the sum of the squares of all distances. The minimum of this sum was found according to the rules of differential calculus. To determine
k and
b, the following equations were obtained:
If we consider that the line passes through the origin, then the linear regression is:
To find the minimum of the equation, the partial values of its derivatives are equated to zero and a system of linear equations is obtained:
The results of bearing capacity of the piles by SLT are represented by blue dots in
Figure 11 and
Figure 12. The values obtained according to
, where
kx is equal to 0.911 and 0.723, respectively, are represented by crosses. The trend lines and corresponding formulas are also presented. As can be seen from the comparative diagram, all points are located below the diagonal, which indicates that all the values of the bearing capacities of the piles after wetting are less than the values of the bearing capacities of the piles without wetting.
From the obtained linear dependencies, linear correlation coefficients were determined, reflecting the degree of closeness of the relationship between two variables. In our case, the correlation coefficient is determined by the following formula:
The correlation coefficient values change in the range: . If the value of the coefficient satisfies the condition , then the relationship between the two variables is weak; if , the relationship is average; if , the relationship is close.
In the case of piles with a protective shell, the obtained correlation coefficient for a straight line not crossing the origin is equal to 0.039, which indicates a weak connection between the function and the variables. For a straight line passing through the origin, the correlation coefficient is 0.61, which indicates the average relationship of the function with the variables. In the case of the ordinary pile, a similar pattern appears. Therefore, to determine the average value of the decrease in the bearing capacity because of wetting the near-pile soils, in both cases, we used linear functions that do not intersect the origin. Thus, after wetting an ordinary pile, the bearing capacity on average decreased by (1–0.723) · 100 = 27.7% (19–38%, see
Table 4), while a pile with a protective and load-bearing shell decreased by (1–0.911) · 100 = 8.9% (4–14%, see
Table 3).
4. Numerical Modeling of a Bored Pile with a Protective Shell in the Plaxis 2D Discussion
The modeling of a bored pile with a protective shell was carried out in an axisymmetric formulation of the two-dimensional Coulomb–Mohr model. The dimensions of the geometric model were taken on the condition that the stress distribution would be negligible within a given zone.
Figure 13 shows a geometric model for the numerical modeling of a bored pile with a protective shell, including a geotechnical element of the site, a pile model with a shell, and a linearly distributed force for a static loading pile.
A simulation was carried out for static tests of a 4 m pile with a protective shell. The pile and the shell represent a single conditional foundation, so there was no need to designate an interface between the pile and the shell; in other words, the possibility of the pile slipping through the protective shell was excluded. According to laboratory and field studies of the soil massif adjacent to the pile, at this construction site, the thickness of the protective shell when the sodium silicate is fed under a pressure of 5–6 atm, on average, is 0.95–1.15 d, where d is the pile diameter. In the numerical modeling, the average value of the containment thickness was taken to be 1.05 d = 0.42 m.
The boundary conditions of the model walls were set in the form of hinged-movable supports with free displacement along the y-axis and displacement along the x-axis = 0. The base of the model was set as a solid embedment, displacement along the x-axis = 0 and y-axis = 0. In the Plaxis software package, these types of boundary conditions were set automatically, as it is suitable for solving most geotechnical problems. The finite element mesh was generated automatically by the program and was a system of triangles. The construction method was based on the stable triangulation principle, which helped to find the optimal mesh sizes.
Before the start of the calculation, the initial conditions were determined, which included the initial geometric structure of groundwater and the initial state of effective stresses. The first computational stage included modeling of the natural stresses caused by gravitational forces; in the second stage—a uniformly distributed load, emitting a static load test loading, was applied to the pile model (
Figure 13).
The main design parameters of geotechnical elements for the Coulomb–Mohr model were taken based on the geotechnical surveys described in
Table 1. The parameters of the containment shell, presented in the form of soil hardened with sodium silicate, were taken based on numerous laboratory tests of hardened soil samples. The physical and mechanical values of the soil, shell, and pile accepted for modeling are presented in
Table 5.
5. Discussion
The protective silicate shell of the new design is technologically different from the previously known methods of ground improvement, but it was also more accurate from the distribution of sodium silicate solution around and beneath the pile, as well as posed better suffusion- and water-impermeable insolation. The results of in situ static loading tests showed that the bearing capacity of the piles casted in protective silicate shells was exceeded by 2.5 times compared to the bearing capacity of the ordinary bored piles without preliminary wetting of the soils and 3.2 times after prolonged soil wetting. Analyzing the loading–settlement curves from six experimental zones in
Figure 14, we can conclude that the loading–settlement curve obtained by numerical modeling had a close relationship with the average experimental curve obtained when testing six SLT with the protective shell. The graph also showed the polynomial characteristics of the obtained curves, with correlation coefficients of 0.999 for both curves. The polynomial curve comprised possible transversely increasing and decreasing extreme pikes of the dependence. In this case, the least squares fit was performed using the following formula:
where
b,
c1…
cn are polynomial function constants.
Using polynomial equations, the settlement was predicted with a further loading pile, both for the experimental and calculated curves. The effectiveness of the proposed technology for the installation of bored piles in a protective and load-bearing shell was given in comparison with the construction technology of the ordinary pile.
As it was determined, the bearing capacity of piles with a protective and bearing shell exceeded the bearing capacity of the ordinary pile by 2.5 times on average. According to this ratio, the required number of ordinary piles and the substitutive amount of piles with a protective and load-bearing shell might be found. Thus, five ordinary piles would respond to the same bearing capacity as two piles with a protective shell. In our example, 645 · 5 = 3225 ≈ 1646 · 2 = 3292 kN. The total cost of the materials and labor costs for the construction of the piles is presented in
Table 6, as well as the economic efficiency, based on modern prices for the installation of bored piles in Kazakhstan.
The advantages of the construction of a bored pile with a protective shell, built by impregnating the near-pile soils with silicate solution, can be considered its environmental friendliness, irreversible corrosion resistance, water resistance, and optimal distribution of strength along the entire length of the shell and in the near-pile zone.
The calculation of economic efficiency of the piles with a protective and bearing shell in
Table 6 confirmed the results obtained according to the functional relationship
to
and expressed by the
kx coefficient. Thus, it was finally obtained that with prolonged wetting of a traditional pile, the bearing capacity decreased on average by 28% (19–38%), and in the case of a pile with a protective and bearing shell, the bearing capacity decreased by 8–9% (4–14%) after wetting.