Influence of mineral composition, surface films and temperature of freezing on dispersity of model sandy soils due to cyclical freeze-thaw

Manukhin Il'ya Vladimirovich Engineer of the 2nd category; Lomonosov Moscow State University, Faculty of Geology, Department of Engineering and Ecological Geology 119324, Russia, Moscow, Leninskie gory str., 1, of. CV-01 il.hrommann@gmail.com

Nikolayeva Svetlana Kazimirovna PhD in Geology and Mineralogy Associate Professor; Faculty of Geology, Department of Engineering and Ecological Geology; Lomonosov Moscow State University 119324, Russia, Moscow, Leninskie gory str., 1, of. CV-01 nikolaeva.sk@gmail.com

Introduction

In connection with the development of Arctic territories, the question of the construction of structures, including linear ones, on bulk soils, which are represented by dispersed incoherent soils – sands, is increasingly being raised. These soils, being in the zone of excessive moisture ^[1] and numerous phase transitions of water ^[2], can undergo significant changes that must be taken into account for the safe and long-term operation of engineering structures.

Cyclic freezing-thawing (hereinafter referred to as CPO) in natural conditions has a very complex and multifaceted effect on soils, which can lead to changes in the structure, condition and properties of soils. In sandy soils, the effect of CPO is most often reduced to the crushing of particles of sandy and larger dimensions and a change in porosity as a result of the formation of cryogenic textures.

The aim of the study was to identify patterns of changes in the dispersion of monomineral sandy soils in CPO, depending on their mineral composition, the presence of autigenic films and the freezing temperature.

The object of the study were samples of monomineral sandy soils subject to cyclic (multiple) freezing–thawing.

The subject of this study is the knowledge about the effect of CPO on the dispersion of sandy soils.

The crushing of sand particles during CPR is widely known, and has been described by many researchers in the last century. Thus, the main mechanism of destruction of sand grains was clarified, which consists in their splitting during freezing of thin films of water in defects of mineral particles ^[3,4,5]. Chernyakhovsky A. G. believed that the wedging action of ice can only lead to a rough fragmentation of the rock along the planes of stratification and contribute to the accumulation of only coarse-grained eluvium (according to ^[6]). Some researchers believed that the destruction of soils during freezing can only occur by stratification due to the freezing of water in the pores (see ^[6]). In the literature, much attention is paid to the destruction of soil particles by the hydration mechanism as a result of the occurrence of wedging pressure of thin films of adsorbed water, mainly shales, siltstones and clay–carbonate rocks are destroyed in this way ^[7]. Sand particles are almost not destroyed by this mechanism ^[8,9].

Water in thawed dispersed soils can be found in the pores between individual soil particles, as well as in defects of mineral grains. Energetically, in dispersed soils it can be free, with a specific binding energy equal to zero, osmotic and capillary (with a low specific binding energy up to 0.53 kJ/mol for capillaries up to 10-7mm in diameter ^[10], adsorbed with the highest binding energy up to 120 kJ/mol ^[11]. It is widely known that bound (adsorbed) water has abnormal properties compared to free water, namely, increased density, increased viscosity, reduced solubility, reduced permittivity and, most importantly in the framework of this work, a reduced freezing point ^[12].

Crushing of sand particles during cyclic freezing—thawing occurs as a result of an increase in the volume of water during freezing by 9%, resulting in pressure, which, when a certain value is reached, can lead to a particle rupture along the surface of a defect filled with water. The value of this pressure is theoretically 13.4 MPa/°From ^[4]. According to other calculated data, the pressure exerted by water during freezing can reach 610 atm (61.8 MPa) at -5 °C, 1130 atm (114.5 MPa) at -10 °C and 2115 atm (214.3 MPa) at -20 °C ^[13]; in experimental studies on freezing water droplets on a free surface, the developing freezing pressure did not exceed 0.02 MPa ^[14]. The freezing pressure observed in dispersed soils turns out to be less than the theoretical values described above and ranges from 1-2 MPa and up to 25 MPa, depending on salinity and degree of moisture saturation in the freezing temperature range from -1 °C to -10 °C under conditions of a gradient temperature field (with the possibility of moving pore water) ^[15,16]. However, high pressures comparable to theoretical values are likely to occur in fine defects of sand grains.

Sandy soil particles are mostly represented by fragments of mineral crystals, most often primary silicates, sometimes rock fragments. Such particles often have certain structural defects, such as microcracks, chips, gouges, etc. Water, when interacting with the surface of such debris, can fall into these defects and be held there by surface tension, gravity, and other interactions. Due to the low energy of the relationship of water with the surface of such minerals, the water in them must freeze at temperatures close to freezing in a free state, however, in the absence of free space for expansion, this does not happen, and the water is in a supercooled state until the pressure exerted by the freezing water exceeds the strength of the limiting circuit.

When conducting studies of the crushing of sand particles in dispersed clay soils, it turned out that sand particles can be crushed even in a matrix of clay particles ^{[17, 18]}. Such a process in natural conditions is very common in the cover sediments of the areas of development of seasonally frozen and thawed soils, this effect is best seen in the cover sediments of watershed massifs, where the influence of transport processes is minimal; there the number of sand particles increases along the section with depth, and the content of dusty particles decreases naturally within the zone of seasonal freezing/thawing ^[19].

In connection with the above material, it becomes clear that during cyclic freezing-thawing of sands, an increase in their dispersion (specific surface area) and a change in their granulometric composition should occur, on which the properties of sandy soils significantly depend.

Objects and methods of research

For the experiment on cyclic freezing-thawing, samples of natural and model sands (including those cleared of surface films) were taken, the description of which is presented below.

The sand of fluvioglacial genesis (fIIms) selected in the Moscow region was used, brown, hardened, consisting of 95% quartz, 4% feldspar, 1% amphiboles, on the basis of which it is considered as essentially quartz (sample K1). From this sand, by dry sieving (without prior dispersion) through sieves with a hole diameter of 0.5 mm and 0.25 mm, the most representative part was isolated – a medium sand fraction of natural dispersion (sample KF1).

To study the effect of natural films on the possibility of crushing sand grains at CPO, their differences devoid of films (samples K2 and KF2) were prepared from the samples described above. Purification from films was carried out by repeatedly heating the sample to 40-50℃ for 20-30 minutes in a 5% hydrochloric acid solution with further rinsing with water (decantation) to a pure filler liquid (with a negative reaction to chlorine ion), after which the latter was drained, and particles smaller than 0.05 mm were extracted from the remaining sediment ^[20]. When treated with a 5% hydrochloric acid solution, most of the substance of the films, with the exception of aluminum oxide, which almost does not react with hydrochloric acid, forms highly soluble compounds that are removed during decantation and drying.

In order to study the effect of the mineral composition of sand grains on their stability at CPO, feldspar and carbonate (calcite) fine sands (samples of PS and Kb) obtained by crushing minerals (except quartz) were taken.

Cyclic freezing-thawing for all samples was carried out for 120 cycles according to the scheme: 4 hours in the freezer and from 4 to 12 hours at room temperature (~25 °C). In natural conditions, soils can undergo from 1-2 cycles of freezing-thawing per year in the flesh to 100 or more under certain conditions (in very cold climates, thin snow cover, insolation and shallow depth (up to 10 cm) ^[21], while on average the soils of the active layer undergo 4-5 cycles per year ^[2,21,22]. Thus, the data obtained by us for 120 cycles of the CPE will describe the transformation of the granulometric composition of sand samples after 24-30 years, which is comparable to the timing of major repairs of the roadbed of road structures (GOST R 58861-2020).

Each sample was placed in a metal bux, brought to a dense addition and complete water saturation, after which it was wrapped with a stretch film to prevent evaporation and sublimation of water (in thawed and frozen states, respectively), thus, the CPR occurred in a conditionally closed system. Quartz samples were synchronously frozen at temperatures -20°C and -10°C, and PS and KP samples only at -20°C. The temperature values were chosen to ensure the freezing of bound water (if any) in the absence of hydrophilic clay minerals in the samples.

The determination of the granulometric composition of sand samples, initial and after the completion of the experiment, was carried out by the sieve method in accordance with GOST 12536 – 2014 with double repetition (Table. 1), the names of the soils are given in accordance with (GOST 25100-2020). Based on the results obtained, the following were calculated: the coefficient of heterogeneity according to A. Hazen Cu=(d60/d10); the coefficient of sorting according to P.I. Fadeev Kof=(d90/d10); the coefficient of sorting according to P.D. Trask Kot=(d75/d25); the coefficient of heterogeneity according to V.D. Melentyev N=d50(d90/d10); the parameter of maximum inhomogeneity (according to I.V. Dudler) Pm=d50(d95/d5), where dx is the particle diameter, less than which the sample contains x% of particles ^[23].

Additionally, indicators of relative stability of Ncr = (a-b)/(a*n) were calculated, where a is the amount of granulometric monomineral fraction in the sample before exposure to cryogenic factors; b is the amount of the same fraction after exposure; n is the number of cycles of freezing and thawing; and for some samples — an indicator of intensity variability of the Kism dispersion ^[24]. In addition, the authors have proposed a new indicator — the coefficient of cryogenic crushing of CCI, which is described in the next section.

The well-known coefficient of cryogenic contrast (KCC) is not considered in this work, since it is intended to determine the cryogenic genesis of sediments ^[25,26].

The morphology of the particles of individual fractions isolated during granulometric analysis was studied using a Levenhuk DTX 500 optical digital microscope.

Results

Theoretically, the crushing of particles at the soil TCO occurs in all sand fractions, while the crushing products may have different sizes, which greatly complicates the interpretation of the obtained data of the granulometric composition, since the loss of mass of one fraction can be compensated by crushing larger particles (see Table 1). Moreover, the interpretation turns out to be very difficult, especially for bi- and polydisperse soils, therefore, for simplicity and clarity of the description of the results, it is convenient to use coefficients traditionally used to characterize the granulometric composition of sands.

Table 1. Granulometric composition of the studied sand samples before and after cyclic freezing-thawing at freezing temperatures of -10 and -20℃

Note: the content of particles by fractions before/after the CPO is given.

It follows from the obtained data of the granulometric composition that the greatest changes as a result of repeated freezing-thawing occur in the most representative medium and fine fractions of sand samples, due to the crushing of which the number of particles of a smaller size increases – fine sand and dusty.

Table 2 shows the values of the average particle diameters (d50) in the studied sands before and after the CPO, which follows: 1) the average diameters decrease slightly or do not change, which indicates a different degree of particle crushing; 2) the presence of surface films increases the intensity of grain crushing on the example of quartz sands; 3) the effect of freezing temperature (at selected values) on the value of d50 is not significant. By reducing the intensity of cryogenic crushing, the studied samples are arranged in the following row according to their mineral composition: carbonate sand, feldspar sand, quartz sand (without films), which is explained by the greatest stability of quartz in the hypergenesis zone compared with other minerals.

Table 2. Change in the average particle diameter of the studied sands as a result of

120 cycles of freezing-thawing

Note: samples K2, KF2 without natural surface films.

According to the results of the experiment, various indicators of the degree of granulometric heterogeneity of sands were calculated (Table 3). These coefficients were proposed at different times by well-known researchers of sandy soils (as well as the corresponding classifications); differing in details, they all follow the general principle: the higher the value obtained, the more heterogeneous the sand is in dispersion.

According to the most well-known indicator of the heterogeneity of Si, it can be seen that almost all quartz sands (K and CF samples) were initially very homogeneous (Si <3), but as a result of cryogenic crushing, their degree of heterogeneity increased, with the exception of the K2 sample frozen at -20 ° C, in which there was no crushing of the predominant fractions, and differences in the granulometric composition were manifested due to the heterogeneity of the sand sample. Feldspar sand in its initial state was heterogeneous in terms of Si, and after CPO it became more homogeneous in all respects, as a result of an increase in the content of fine sand fraction, which eventually became predominant. In carbonate sand, in almost all indicators, the degree of heterogeneity has hardly changed due to the fact that the fragments formed as a result of crushing turned out to be predominantly the size of the predominant fraction (see Table 1).

Table 3. Changes in the values of various indicators of heterogeneity of the studied sands as a result of cyclic freezing-thawing

Notes: 1) in each cell, the values of the coefficients before/after the CPR are shown; 2) sand samples in which the degree of heterogeneity (according to a certain indicator) increases are marked in green, and in which it decreases in red; the absence of color means that there are no changes or they are minimal (within accuracy).

It seems important to compare the data we have obtained with the results of previous studies. Thus, V. N. Konishchev in 1981 in the monograph "Formation of the composition of dispersed rocks in the cryolithosphere" gave a list of published experiments on cyclic freezing-thawing of monomineral monofractive sands, for which the relative stability index of the Nkr was calculated; the lower its value, the higher the stability of the fraction.

This indicator is intended to describe cryogenic grinding in monomineral monodisperse sands, in the studied samples no sample strictly fits its definition, since even the most homogeneous and monomineral of them, samples K2 and K22, contain up to 23% particles of fractions other than the dominant 0.5-0.25 mm. Although our samples and the conducted experiment do not fully comply with the conditions described in the literature ^[3], the results obtained can be used for an estimated comparison.

Table 4 shows the published data on the experimental TCO of monomineral monodisperse sands and the relative stability indicator calculated for them. The following patterns are immediately visible here: an increase in resistance to cryogenic grinding with a decrease in dispersion and its decrease with a decrease in freezing temperature ^[3]. The data we have obtained generally fit into the previously identified patterns, however, the values of the relative stability of minerals differ from those previously published at times. Lines 1-2 and 4 of this table attract attention, they show the results for quartz sands that do not have a significant number of autigenic films: for samples in lines 1 and 2 – by obtaining sands by crushing from a large monomineral crystalline sample; 4 – by chemical removal of films. At the same time, the last sand sample showed significantly smaller crushing scales, most likely due to fewer grain defects, compared with artificially ground minerals.

Table 4. Comparative table of the results of calculating the indicator of relative stability of minerals according to new and published data (according to ^{[3, 6, 8]}).

Note: all samples were in a water-saturated state.

When considering the crushing of quartz sands at a freezing temperature of - 20 °C (lines 5-11, in which the samples are essentially represented by 4 natural sands divided into fractions and natural sand divided into fractions and purified from films), we see that the relative stability of minerals according to published experiments turns out to be relatively close to each other, at the same time, for the sample of CF1, it turns out to be several times less, and for CF2, the indicator is generally zero, which confirms the negative effect of natural films on the stability of grains to cryogenic crushing.

The stability of feldspars turns out to be slightly higher on average than that of carbonates, which is consistent with the data given earlier ^[3].

During the analysis of the obtained heterogeneous data, it was decided to find an indicator (coefficient) that would describe the transformation of the entire granulometric (fractional) composition of soils as a result of cyclic freezing-thawing. For this purpose, a mathematical interpretation of the course of cryogenic crushing was carried out (Fig. 1), it shows that as a result of the CPR, the content of the Ci fraction decreases by the value of fi and the redistribution of the crushed material into smaller fractions in the proportion xij.

Fig. 1. The scheme of transformation of the granulometric composition of the soil as a result of cyclic freezing-thawing, where Ci is the content of the I–th fraction of the soil before the CPO (C1, C2, ..., Sp is the content of the entire range of standard fractions from the largest to the smallest), %; ci is the content of the i-th fraction of the soil after the CPO, %; fi – relative mass loss in fraction i, due to cryogenic crushing, %; xij – coefficient showing which fraction of fi was crushed to fraction j, d. units.

It is quite obvious that in order to obtain the effect of cyclic freezing-thawing on the granulometric composition of the initial soil, we need to find out the sum of the values of fi, for this we need to obtain the modulus of the difference Ci and ci for each fraction, after which we sum up all the obtained modules and divide the sum by 2. The formula for calculating the cryogenic crushing coefficient proposed by the authors is presented in expression (1), where T is the number of freeze-thaw cycles.

. (1)

From the definition (1) of this coefficient, it follows that due to the use of the difference module, it will give a sign–neutral characteristic of the change, that is, an increase and decrease in dispersion by the same amount will give the same value, although completely different processes will be described. It shows the relative content of particle fragments formed as a result of one freezing-thawing cycle in the experiment.

In fact, this coefficient clarifies the general indicator of the intensity of variability of the Kism dispersion, which was used in the monograph "Microstructure of frozen rocks" (1988). It differs from our coefficient in that it was created to assess the intensity of variability of the granulometric composition in general, and not only within the framework of the crushing assessment, moreover, it "doubled" the results of changes in the granulometric composition by taking into account the formed fragments from a larger fraction and their accumulation in smaller fractions.

In fact, our proposed coefficient of cryogenic crushing of CCI shows the number (%) of debris formed during 1 freezing-thawing cycle. Knowing its value, the approximate number of freezing cycles per year and the service life of an earthen structure, it is possible to predict and evaluate the change in the granulometric composition of soils.

The data of the CCI calculations (Table. 5) confirm the previously published conclusions. A decrease in the freezing temperature on average leads to a higher crushing intensity (higher values of the KCI coefficient are obtained), while the proportions in which it changes very often do not coincide with the expected increase in theoretical pressure values during water crystallization of 100-200 MPa (-10 — -20 °C) ^[13].

The influence of surface films on quartz grains was quite clearly manifested, especially in comparison with the published data. Thus, sands with grains peeled from films show mainly a value of KCI = 0.025%, while sands with films show values from 0.039 to 0.630% in samples with similar dispersion, with the most common values being about 26 * 10-2. The minimum value in this range (3.9*10-2) belongs to Lyubertsy sand, which has passed 1000 cycles of freezing-thawing, and its granulometric composition underwent most of the changes (~ 90%) after 150 cycles, and then it hardly changed, as a result of which the final value turned out to be "underestimated".

The reason for the reduced cryogenic pulverizability of quartz purified from films is related to their water-retaining ability. The quartz sand taken for research had a brown color during macro description, which usually indicates a significantly ferruginous composition of surface films (or "jackets"), represented by a complex complex consisting of Fe, Mn and Al hydroxides firmly bound to the grain surface ^[19,27,28].

Table 5. The results of calculations of the KCI cryogenic grinding index based on the obtained new and published data (according to ^[3,6,8,24]).

Notes: 1) the initial state of the samples was water–saturated; 2) for soils with bi- and polydisperse granulometric composition, the resulting coefficient was placed in a pair of cells responsible for the predominant fractions.

In addition, the surface of quartz particles in natural conditions in the presence of water is covered with a thin layer of amorphous silica or a layer of disturbed crystals up to 100 Å thick. This layer has higher values of specific surface area (from 300 to 900 m2/g) and adsorption capacity, it is formed as a result of partial surface dissolution of quartz in water or precipitation of dissolved silica from water ^[29,30]. The process of dissolving quartz in water is very complicated, since quartz itself is dissolved here, as well as amorphous silica, while in parallel amorphous silica is often deposited from the solution onto the active surface of quartz, slowing its further dissolution. The solubility of silica in water decreases due to the presence of metal ions in water, including Fe and Al as a result of their chemisorption on the silica surface ^[29].

Films on grains of K1 and KP1 sands are mainly represented by Fe, Al hydroxides and amorphous silica, the first 2 components were removed using a 5% HCl solution. The treatment used was not enough to completely remove the films, however, it is clearly visible that the number of colored particles decreased significantly (Fig. 2). As a result, the surface of the sand grains in the K2 and KF 2 samples (after purification from the films) lost most of the Al and Fe hydroxides, and mainly amorphous silica remained on it, in the samples On the other hand, K1 and CF1 hydroxides of Al and Fe are adsorbed on the surface of amorphous silica, and first of all they are adsorbed on the "porous" surface of amorphous silica, due to the higher heat of adsorption compared with a flat surface ^[30]. Thus, in both cases, our grains are covered with thin films of highly dispersed and finely porous amorphous silica, but in the case of samples K1 and CF1, Al and Fe hydroxides are adsorbed on the pore walls in amorphous silica, which, due to their location, make the mass of the amorphous substance even more finely porous, increasing the water retention capacity.

The effect of surface films on the crushing of quartz sand grains during CPO is most likely due to the ability of the amorphous substance of the films (oxides and hydroxides of iron and aluminum, amorphous silica, etc.) to retain thin films of bound water on its surface. The substance of the films tends to concentrate near defects in the crystal lattice of minerals (cracks or chips), as a result of which the films of bound water are confined to the same areas and, judging by the data obtained, can significantly freeze in sand at temperatures up to -10 ° C. Thus, sands with a larger number of films on the grains are crushed at a higher temperature than their counterparts with partially removed films. There is also an assumption that films of amorphous matter on the surface of particles can play the role of "plugs" at the mouth of defects, causing water to freeze in them.

Fig. 2. Surface character of grains of sands K1 and KF1 with natural films (a) and sands K2 and KF2 with removed films (b)

The influence of surface films in quartz sand also extends to the nature of the dispersion change. Thus, in quartz with natural surface films, from 1 to 4% (3-4% prevail) of dusty particles (less than 0.05 mm in size) were formed, while in quartz sand cleaned of films, from 0 to 4%, and mainly 0-1% of dust particles were formed.

The practical application of CCI is to use it to assess changes in the granulometric composition of sandy soils as a result of a certain number of cycles of freezing and thawing. Knowing the average annual number of soil temperature transitions through 0 ° C and the estimated service life of the soil base or roadbed, it is possible to obtain a quantitative content of newly formed debris particles in the soil during this period. Knowing the patterns of crushing and distribution of its products, it is possible to obtain the distribution of these fragments by granulometric fractions. It is well known that the crushing intensity of quartz grains is inversely proportional to their size ^[3], and it is also clear from the presented materials that during cryogenic crushing of quartz fragments of a "neighboring" smaller fraction are most often formed (see Table. 1), fragments of subsequent small fractions are formed as a rule less frequently, and the smaller the fraction, the less likely such fragments are to form. This rule works up to and including the fine sand fraction, coarse-grained particles can form when crushing particles of different sizes with a significant probability (Table 6).

Table 6. Results of the analysis of the distribution of debris by fractions

and its changes during cryogenic crushing

Notes: 1) the average granulometric composition of the samples before/ after CPO is given; 2) the columns Δ show the results of CPO exposure: with the sign "-" – the mass fraction of formed fragments (broken off) from this fraction; with the sign "+" – the mass fraction of fragments broken off from larger fractions (percentages of formed fragments are given which fell into this fraction); 3) the sign * indicates changes in the content of fractions, taking into account the heterogeneity of the samples.

When choosing such a technique, the question arises about the selection of cryogenic grinding coefficients used for calculations. The researchers obtained different data for different minerals, of different sizes at different temperature conditions, at different soil humidity, and for experiments of different lengths (the number of freezing—thawing cycles). The model conditions of the selected coefficient should be as close as possible to the simulated conditions in the calculation situation, while, as in many engineering calculations, it is worth adhering to the principle of calculating the worst possible option.

For example, let's present a calculation of the change in the dispersion of medium-sized sand in the roadbed. According to clauses 5.1 and 5.3 of GOST R 58861-2020, major repairs of the roadbed are carried out every 24-30 years, which can be taken as a time interval for calculation. The temperature regime and the average number of freezing-thawing cycles per year should be determined from archival data of the nearby area for soils of similar dispersion. In this situation, an approximate calculation is being carried out, we will take 5 cycles per year and the freezing temperature is -10 ° C, so it turns out that 120-150 freezing-thawing cycles should occur during the calculation period. The soil, within the framework of this estimated calculation, is assumed to be water-saturated, which will correspond to the principle of the possible worst case scenario. We also believe that the sands are predominantly quartz and have a significant number of films on the surface of the grains, so that from our materials (Table. 5) Line 3 with a KCI value of 0.092%/cycle will be best suited.

The granulometric composition of sand and the results of the forecast of its change after 120 cycles of freezing-thawing are shown in Table 7.

The content of newly formed fragments x, for it should be x = KCI*n, where n is the number of freezing–thawing cycles, therefore x =0.092*120, hence x =11%. Further, to predict the granulometric composition, we need to understand how the destruction products are distributed into smaller fractions, for this we will use the analogy method. For the sandy soil taken as an example, the closest analogue is sample K1, the patterns of crushing and distribution of debris by size are shown in Table 6, and they will be used in our forecast.

Table 7. The initial granulometric composition and the results of assessing its change after 120 cycles of freezing-thawing for medium-sized sand

As can be seen from Table 8, the dispersion of these sands according to the results of our forecast should change very significantly, which is especially clearly seen from the values of the Hazen inhomogeneity coefficient (Si), since homogeneous sand has become heterogeneous. To illustrate the effect of such a change in dispersion, we calculated the filtration coefficients for the initial and predicted granulometric compositions of sandy soil according to published formulas ^[23,27]. From the data obtained, it can be seen that the predicted changes in dispersion should lead to a significant decrease in the filtration coefficient (by more than 4 times), therefore, we believe that these transformations of dispersion should not be neglected.

Table 8. Changes in the values of various dispersion indicators and the calculated filtration coefficient of medium-sized sand based on the results of the forecast of the impact of 120 freezing-thawing cycles

Note: the filtration coefficients were calculated using the Hazen ^[27] and Slichter ^[23] formulas, for a soil porosity of 35% at a water temperature of 10 °C.

For greater reliability of the results of using this approach, a database of cryogenic grinding coefficients is needed, which will take into account the influence of a whole range of factors, such as temperature conditions, different soil moisture, their mineral composition (up to different modifications for the most common minerals, for example, quartz), density, the amount of physico-chemical activity (reflects the presence of surface films). Also, a separate issue is the study of the patterns of change in the granulometric composition of polydisperse and polymineral sands.

Conclusions

1. In the conditions of the experiment, cyclic freezing-thawing leads to a change in the granulometric composition of model sandy soils, an overall increase in dispersion due to crushing of part of the grains. Depending on the mineral composition, there is a decrease in the intensity of cryogenic crushing in the series: carbonate sand, feldspar sand, quartz sand (without films on the grains), which is explained by the greatest stability of quartz in the hypergenesis zone compared with other minerals.

2. The resistance of quartz grains to cryogenic weathering is significantly influenced by surface films. The intensity of grain crushing under experimental conditions increased due to an increase in the physico-chemical activity of the grain surface due to the substance of the films.

3. To understand the nature of the change in dispersion as a result of cyclic freezing-thawing, the use of well-known indicators of sand heterogeneity and average particle diameter turned out to be little informative. In most cases, the sands become more heterogeneous due to grain crushing, but the opposite can also be observed, depending on the initial granulometric composition.

4. A new indicator is proposed to describe the transformation of soil dispersion as a result of cyclic freezing-thawing – the coefficient of cryogenic crushing, reflecting the increase in the content of products of destruction of soil particles during a single cycle of freezing-thawing.

5. The higher values of the cryogenic crushing coefficient obtained confirm the previously published conclusions that a decrease in the freezing temperature (at values of -10 ° C, -20 ° C) on average leads to a higher intensity of crushing of soil particles.

6. The use of the coefficient of cryogenic crushing makes it possible to predict changes in the granulometric composition of sandy soil under the influence of cyclic freezing-thawing over a certain period of operation of an earthen structure.

7. The change in the dispersion of sands as a result of cyclic freezing-thawing should not be neglected, since due to the crushing of sand grains, a thinner sandy and coarse-powdered material is formed, which in turn significantly changes the properties of soils.