Soil stabilization by injection of mineral complexes

. Soil stabilization and strengthening is a necessary complex of engineering measures and can be carried out by introducing organic or mineral substances into the soil which increases soil viscosity and its mechanical characteristics. Soil modification can also be accompanied by mechanical pressure, as well as temperature influence. Judging by the results of the analysis of the state of the issue, a hypothesis was formulated, according to which an improvement in the operational characteristics of the soil is possible as a result of the use of a stabilizer based on quicklime and an active mineral additive. The object of the study was light sandy loam. The experiment was carried out using mathematical planning methods with subsequent statistical processing of the results. The following are accepted as variable factors: the consumption of quicklime, the consumption of ground slag and the specific surface area of ground slag. The strength of the modified soil at the age 28 of hardening under laboratory conditions and its average density were taken as response functions. It was established that the introduction of quicklime and finely ground slag could significantly improve the operational characteristics of the soil as a result of analytical optimization of the obtained strength and density functions of the modified soil.


Introduction
Soil, including clay, loam or sandy loam, in an unstable state can not only disrupt technological and logistical processes, but also pose an immediate danger to human life (Fig.  1). Soil stabilization and strengthening is a complex of engineering measures aimed at improving reliability, durability and mechanical characteristics of soils as bases for buildings and structures. These impacts are based on the modification of soil properties by mechanical and chemical methods [1][2][3]. All types of soils and stabilizers, as well as the technological design of the processes must be compatible. The following groups of characteristics of natural soils have the most significant impact on the effectiveness of soil stabilization. Firstly, it is the content of clay or (and) sand particles. Secondly, it is the plasticity number and the hydrogen index of the soil. Thirdly, it is the content of humus substances, gypsum, chlorides and sulfates. A very important factor is the hydrogeological condition of the soil, including its natural moistening, the groundwater level, the level of flood waters, as well as the possible presence of underground rivers, karst voids, etc. Therefore, when assigning one or another method (technology) of stabilization, studies of soil properties should be carried out in an obligatory way [4][5][6].
Soil stabilization is carried out in order to change the physical and mechanical properties of this soil. Stabilization is carried out by introducing a mineral binder, synthetic or organic emulsions, mineral additives in combination (or without it) with thermal impact or mechanical sealing.
Sandy soils, loams and sandy loams can serve as stabilizers by distributing and averaging corrective additional soil; this can be achieved by the introduction of a stabilizer into the soil of natural density. According to the effect produced on the soil material, stabilizers are divided into water repellents and hardeners. Firstly, hardeners do not only reduce the development of heaving processes during soil freezing, but, secondly, they significantly change the water-physical and physical-mechanical properties of the soil [7][8][9].
The general classification of soil stabilizers is based on compliance with the following features: the state and properties of the object, as well as the specifics of the implementation of technology, taking into account the characteristics of the object. Both mineral and organic materials can be used as stabilizers: bituminous, lignins and lignosulfonates, vinyl-acrylic copolymers, as a rule, in combination with a mineral component.
The scientific foundations and basic theoretical principles for the stabilization of dispersed systems, including soils, were laid down in the works of Academician P.A. Rebinder and his team in the framework of creating the foundations and practical applications of physical and chemical mechanics. The introduction of chemicals into the soil, in addition to direct reactions aimed at strengthening the soil, makes it possible to change the surface potential of soil particles, which improves the process of its compaction [10,11].
As a result of stabilization, the soil structure changes at the micro level (Fig. 2). This effect is achieved due to the growth of crystals as a result of the reaction between calcium (lime) ions SiO2 and Al2O3 in the presence of an additive with pozzolanic activity binding an excessive amount of CaO. Modern methods of soil stabilization are based, among other things, on the use of inorganic binders, both air and hydraulic hardening, as well as mineral additives having both pozzolanic and hydraulic activity. Depending on the type of soil and the required level of stabilization, mineral or organomineral complex additives can be used. There are systems that use Portland cement, Portland slag cement, pozzolanic cement, lime, slag binder, fly ash, gypsum, ground ferrous metal slag. Soils stabilized with slag binder and mechanically compacted are a new building material [12][13][14].
The preliminary studies have shown that a complex stabilizer, including quicklime and finely ground blast-furnace slag, can significantly reduce the swelling index of clay soil. In accordance with the review carried out, a hypothesis was formulated which declares a possible improvement in the operational characteristics of the soil as a result of the use of a stabilizer based on quicklime and an active mineral additive.

Materials and methods
Quicklime GOST 9179-2018 ("Construction lime. Technical conditions"). Lime crushed to residues on sieves No. 02 and No. 008 no more than 1.5% and 15%, respectively. The content of active CaO and MgO 80%; unextinguished grains -no more than 8%. The ultimate strength of the samples in compression after 28 days of hardening 2.0 MPa.
Light sandy loam was used as a model soil ( Table 1). The natural moisture content of the loam was 19.6%; humidity at the yield point was 32.6%; humidity at the rolling point was 24.1%. The loam plasticity number was 9.1; the maximum density was 1740 kg/m 3 and the optimum moisture content was 18.6%. The granulometric composition of loam is presented in the table 1. subsequent formation of regression equations, testing of statistical hypotheses and assessment of the adequacy of the obtained models with subsequent interpretation: a graphical one and a computer based one. The experiment was carried out according to the full factorial D-optimal and rotatable plan.
The optimization of the obtained equations was carried out by an analytical method to determine the local maximum of mathematical polynomials as functions of several variables. This methodological technique was developed at NRU MSCU and received the name of "analytical optimization" and is widely used in the analysis of the results of mathematical planning of the experiment [15][16][17].
As variable factors, the following are accepted: the consumption of quicklime (Х1) in % by weight of the soil, the consumption of ground slag (Х2) and the specific surface area of ground slag (Х3). Experimental conditions are given in Table 2. The strength of the modified soil (U1) at the age of 28 hardening in laboratory conditions and its average density (U2) were taken as response functions.

Results and their discussion
Statistical processing of the results of the active experiment made it possible to obtain regression equations for compressive strength (U1) and average density (U2). The significance of the coefficients was assessed by comparison with the confidence interval Δb, calculated by Student's t-test (t-test). The values of confidence intervals for strength and average density were respectively Δb1 = 0.3 MPa, and Δb2 = 6 kg/m 3 . Coefficients smaller in modulus of confidence intervals were considered insignificant and equated to zero. As a result, the following mathematical models (polynomials) were obtained: -for compressive strength: -for medium density: The analysis of the obtained regression equations allows us to draw the conclusions about the influence of variable factors on the result. The compressive strength (U1) is most affected by the consumption of quicklime (the coefficient at X1 is 1.4). The influence of the degree of grinding (specific surface area) of ground slag is manifested to a lesser extent (the coefficient at X3 is equal to 0.6) and has an extreme character (the coefficient at X3 2 is equal to -0.5), which makes it possible to optimize the dependences analytically (1 and 2). The combined effect of the consumption of quicklime and ground slag is also significant (the coefficient at X1X2 is 0.6). This joint influence has a pronounced synergistic character. Then we carry out the analytical optimization of dependencies (1 and 2). To do this, we determine the partial derivative of the function Y1 with respect to X3 and find the optimal value of the specific surface and then solve equations (1 and 2) at the optimal value of the specific surface.
We solve dependences (1 and 2) at Х3=0,6: -for compressive strength: Y1 = 6,1 + 1,4Х1 + 1,0Х2 + 0,6×(0,6)+ 0,6Х1Х2 -0,5×(0,6) 2 -for medium density: We get the following X3-optimized mathematical models (polynomials): -for compressive strength: -for medium density: We determine the naturally optimal value of the specific surface area of ground slag: Ssl = 300 + 100×(0,6) = 360 m 2 /kg The graphical interpretation of dependences (3 and 4) is presented in the form of a nomogram (Fig. 2). Sector 1 is for interpreting the average density functions; sector 2 is aimed at interpreting the compressive strength function. The solution of the prognostic problem is as follows: the experimenter sets the values of the variable factors and, then, based on these values, the compressive strength of the modified soil and its average density are determined. In Fig. 3. an example of solving a prognostic problem is shown. We set the values of lime consumption as 9% and slag consumption as 15%. From point a we draw a straight line (a,) parallel to the x-axis. From point c we draw a perpendicular to the intersection with a straight line (a) and we get the value of the average density of 1865 kg/m 3 . Then from the point we draw a perpendicular to the intersection with the straight line (a) and we get the value of the soil strength equal to 7.0 MPa. Thus, with a lime consumption of 9% and a slag consumption with a specific surface area of 360 m 2 /kg, we obtain soil (modified loam) with a density of 1865 kg/m 3 and a compressive strength of 7.0 MPa. Next, control moldings are carried out and, as a result of an active experiment, the real indicators of samples of modified light sandy loam are determined.

Conclusions
Soil stabilization is a necessary engineering measure, and is especially effective in the case of weak soils, as well as soils with decreasing cohesion as a result of moisture.
The advantages of soil stabilization and strengthening technologies lie in the possibility of implementing the following solutions. Firstly, this is a reduction in the volume of excavation of soft soils. Secondly, they lead to the minimization of the amount of imported inert materials. Thirdly, high productivity of works on strengthening the soil is achieved. Fourth, we get the possibility of forming a base that is not susceptible to frost heaving. Fifthly, it is a continuous process of gaining strength by stabilized soil.
It has been established that the use of a complex modifier based on quicklime and ground blast-furnace slag makes it possible to obtain a modified soil which is quite suitable for use as the foundation of construction objects.