DETERMINATION OF THE SUBSIDENCE OF STRUCTURES ON LOESS-LIKE LOAMS OF KUZBASS

F. P. NIFANTOV

(Presented by the scientific conference of the Geological Prospecting Faculty)

Introduction

Surficial loess-like deposits in many regions of our country rapidly disintegrate when soaked with water and cause uneven settlements of the ground surface and of structures. This specific property, inherent only in macroporous loess-like soils, is called subsidence.

Many structures built without preliminary determination and allowance for the subsidence of loess-like soils subsequently experienced large, rapidly developing uneven settlements and deformations.

In Kuzbass the first subsidences of loess-like loams were recorded in 1931–1933, when certain large structures, upon soaking of the soils, settled rather quickly and unevenly, in places by 50–80 cm, and, as a result, underwent substantial deformation (Abelev, 1934). Such settlements and deformations of structures on subsiding soils were observed in the Stalinsk, Kemerovo, Belovo, and other districts of Kuzbass. In some places they are still observed today, where construction is carried out without taking subsidence into account and without applying appropriate measures. In recent years, as a result of the subsidence of loess-like loams, a number of large structures of new mines, workers’ settlements, and cities have been deformed.

The deformations of structures built on loess-like soils in other industrial regions of the USSR led to the development of broad investigations into the properties of these rocks. In the southern regions of the European part of the USSR, investigations were begun as early as 1928–1929. By 1940 a large amount of material had accumulated, partially summarized in the proceedings of two All-Union conferences on questions of construction on subsiding loess-like soils (Collected volume, 1939).

In the postwar period, in connection with the enormous scale of construction in the country, still greater attention was given to the study of the properties and subsidence of loess-like deposits. A number of new works were published, among which the monographs of Yu. M. Abelev (1948) and N. Ya. Denisov (1946, 1953) stand out in particular, giving the most complete treatment of the physicomechanical properties and subsidence of loess-like soils. The method for determining the subsidence of strata of macroporous soils, proposed by Yu. M. Abelev in the aforementioned monograph, has been introduced into the current se—

time standards (Building Standards, 1954). It is recommended that the collapsibility coefficient be determined in a special apparatus by soaking the soil under a load of \(3\ \mathrm{kg/cm^2}\) (Abelev, 1948; Instruction, 1950; Building Standards, 1954).

In the present work, along with the results of laboratory studies of soils, a method is described for determining the magnitude of the actual settlement (additional settlement) of structures, somewhat different from the accepted standard method, developed by the author in 1951 as applied to the loess-like soils of Kuzbass and to actual foundations of structures (Nifantov, 1951).

Some results of laboratory studies of the collapsibility of cover loams of Kuzbass

Taking into account the needs of construction and the widespread occurrence of loess-like loams in Kuzbass, the task was set of studying their collapsibility. Some results of these studies are presented below.

Fig. 1. Graph of compressibility and collapse of macroporous loess-like loam No. 462 in a compression apparatus.
1—soaking and collapse under a load of \(2\ \mathrm{kg/cm^2}\); 2—soaking and collapse under a load of \(3\ \mathrm{kg/cm^2}\); 3—soil moisture content, as percent by weight; 4—porosity coefficient.

The studies were carried out according to the following scheme. From a monolith of loess-like loam with dimensions \(25 \times 25 \times 25\ \mathrm{cm}\), cubes were cut out.

by a volume 3–4 times greater than the internal volume of the working ring of an ordinary compression apparatus, having a height of 15 mm and an area of 40 cm². The cubes were covered with several layers of gauze and placed in dry sand. The samples were kept in the sand at room temperature for several days, then loaded into the rings of compression apparatuses. Each ring with soil, after weighing and installation in the apparatus, was covered with a thick layer of moist cotton wool in order to preserve the soil moisture attained in the sand. Subsequently, loads were applied to the soil in increments of 0.5 kg/cm² and were brought, in the first apparatus, up to 1 kg/cm²; in the second, up to 2 kg/cm²; in the third, up to 3 kg/cm², etc. Under these loads (after stabilization of settlement), the soils were wetted with water from below in order to ensure the best displacement of entrapped air from the soil pores. After settlement ceased, another 2–3 load increments were applied to the soil, and the experiment was concluded.

Fig. 2. Graphs of compressibility and settlement of moist (degree of moisture 0.46–0.66) macroporous loesslike loam No. 713.
1—wetting of the soil before application of load; 2 and 3—wetting under loads of 1 and 3 kg/cm²; 4—soil porosity coefficient.

The tests carried out according to the indicated scheme made it possible to determine the coefficients of macroporosity and settlement \((i)\) of the same soil under different loads. For a number of monoliths the experiments were repeated at other initial moisture contents of the loam. As a result, interesting data were obtained, making it possible to establish the dependence of settlement on the load and the initial moisture content of loams.

The maximum subsidence was shown by macroporous, slightly moist (with a moisture content of 3–8%) loess-like loams having a high total porosity (47–50%). For example, loam 462 (Fig. 1), soaked at a moisture content of 3.3% and under a load of 2 kg/cm², showed a subsidence coefficient of 0.126, and at the same moisture content and a load of 3 kg/cm² an even higher one—0.134. High subsidence coefficients were found for all the other slightly moist macroporous loams as well (see Table 1). But these same macroporous loams, having moisture contents of 15–22% (degree of moisture 0.46–0.66), proved to be less subsiding, especially in those cases where small vertical loads were applied (Figs. 1, 2). Still less subsidence was shown by macroporous loam 679, which had an initial moisture content of 24.3–25% (Fig. 3). Finally, all loams showed no subsidence at all when fully saturated with water before the experiment. In this case ordinary compression curves were obtained

Fig. 3. Graph of compressibility and subsidence of very wet (moisture content 23–25%, degree of moisture 0.67–0.71) macroporous loess-like loam No. 679. 1—soaking of the soil under a load of 0.5 kg/cm²; 2 and 3—soaking and subsidence of the soil under loads of 1 and 3 kg/cm²; 4—soil porosity coefficient.

with high coefficients of compaction, as is evident in the same graphs (Figs. 1, 2, curves shown by solid lines). Similar results were obtained for many other macroporous loess-like loams of the Kuzbass, taken on river terraces and watersheds, and in all experiments the most subsiding proved to be slightly moist macroporous loess-like loams with natural moisture content below 10–12%. With increasing moisture content, subsidence usually decreases, and at degrees

Table 1

Data on the subsidence of some cover loess-like loams of the V terrace of the Tom River

At moisture contents above 0.7–0.8, the same macroporous loams pass into the category of non-collapsible but readily compressible soils. Studies have also shown a change in the degree of collapsibility of one and the same soil depending on the vertical load (see Figs. 1 and 2). With increasing loads, the collapsibility of loams increases, which is apparently explained by a more intensive disturbance of the structure by the hydrodynamic pressure of the water arising inside the pores at the moment of the rapidly occurring settlement of the soil under the additional load. It follows from this that collapsibility should be determined at such optimum moisture contents and vertical loads as will act on the soil during the construction and operation of the structure, and not at a constant load of \(3 \ \text{kg}/\text{cm}^2\) and artificial moisture content, as some instructions (1950) recommend. Only in this case will the magnitude of settlement of the structures being designed be determined with sufficient accuracy.

Method for determining the additional settlement (collapsibility) of structures designed on loams collapsible under load

The relationships established above for the change in collapsibility with initial moisture content and load make it possible to consider that collapsibility in the loess-like loams of the Kuzbass occurs only within the compressible zone of the soil, where additional stresses are created. Outside the stressed zone, soils similar to the loess-like loams of the Kuzbass retain their primary structure, if during soaking no large additional hydrodynamic pressures are created there that disturb the cohesion between soil particles.

Taking the foregoing into account, one may recommend a more accurate method for determining the probable collapsibility (additional settlement) of structures designed on loams collapsible under load. At the construction site, near the outlines of the structures being designed, test pits should be sunk and, within depths corresponding to the depth of the compressible zone, undisturbed samples should be taken every 1.5–2 m. In the case of soils heterogeneous in macro-porosity and moisture content, undisturbed samples must be taken from each layer. From the undisturbed samples, the coefficients of collapsibility \((i)\) of the soil should be determined in ordinary compression devices under such loads as will be transmitted to the soil by the designed structure at the depth from which the undisturbed sample was taken (Fig. 4).

It is evident from the figure that the collapsibility for the first undisturbed sample must be determined under a load of \(2 \ \text{kg}/\text{cm}^2\), and the resulting coefficient of collapsibility \((i_1)\) should be extended to the interval of 1.25 m bounded by the base of the foundation and the line with a stress of \(1.5 \ \text{kg}/\text{cm}^2\). The second undisturbed sample, taken at a depth of 4.5 m, is tested for collapsibility under a load of \(1.0 \ \text{kg}/\text{cm}^2\). The coefficient of collapsibility \((i_2)\) obtained in this case may be extended to an interval of 2.25 m bounded by the stress lines \(1.5\)–\(0.70 \ \text{kg}/\text{cm}^2\). Finally, the coefficient of collapsibility of the loam for the third undisturbed sample \((i_3)\), taken at a depth of 6 m, is determined under a load of \(0.5 \ \text{kg}/\text{cm}^2\) and applies to a layer of 4.0 m bounded by the stress lines \(0.7\)–\(0.32 \ \text{kg}/\text{cm}^2\). The indicated subdivision into intervals, in the case of homogeneous loams, provides the accuracy of calculations necessary for practice. All underlying layers of loam, located outside the compressible zone and not receiving any significant additional stresses, will not be collapsible. In this case, the total settlement of the structure, for the given dimensions of the foundation and a load of \(2 \ \text{kg}/\text{cm}^2\), will be equal to

to the sum of the settlements of the three listed intervals and is calculated by the following formula:

\[ S=i_1h_1+i_2h_2+i_3h_3, \]

where \(S\) is the total settlement of the soil within the compressible zone, in cm;
\(i_1\), \(i_2\), and \(i_3\) are the settlement coefficients of the soil of the first, second, and third intervals;
\(h_1\), \(h_2\), and \(h\) are the thicknesses of the intervals, in cm.

This formula coincides with one of Yu. M. Abelev’s formulas, recommended by him for calculating the partial settlement arising during the infiltration into the soil of the compressible stratum of “a relatively small quantity of water” [1948]. We propose it for determining the total settlement of industrial and civil buildings designed on loams that settle under load and have undergone maximum wetting.

The magnitude of the settlement of buildings, calculated by this formula, is recommended by us for assigning the type and scope of measures designed to preserve the stability of buildings erected on soils that settle under load. Yu. M. Abelev, for this purpose, as is known, recommends calculating the magnitude of settlement of the entire thickness of the macroporous soil at an average settlement coefficient determined under a soil pressure of 3 kg/cm\(^2\), and not under the actual load that will be transmitted to the soil by the structure. It is clear that the actual magnitude of the settlement of buildings, calculated by our method, in the case of thick loams that settle under load, will always be less than the settlement of the stratum determined by Yu. M. Abelev’s method. Thus, the method we propose will ensure savings in funds spent on measures against settlement of structures. The difference in these results is quite clearly emphasized when analyzing a specific example.

Fig. 4. Curves of the distribution of additional and overburden stresses in the soil along the \(z\)-axis beneath the base of a strip footing.
1—curve of additional stresses; 2—curve of overburden stresses; 3—depth of propagation of the compressible soil zone.

On a homogeneous loam that settles under load, with a thickness of 15 m, the construction of buildings was designed with foundations 2.0 m wide, embedded to a depth of 2.10 m and transmitting an additional load to the soil of 2 kg/cm\(^2\). For these foundations and soil, the thickness of the compressible zone, according to the diagram shown in Figure 4, is 7.5 m. The settlement coefficient of the loam \((i_1)\) for the first monolith, selected...

to it at a depth of 2.5 m, when soaking the soil under a load of 2 kg/cm², proved to be 0.06; for the second, taken at a depth of 4.5 m and tested for subsidence at \(P = 1.0\) kg/cm², it proved to be 0.037; and finally, for the third monolith from a depth of 6.0 m, tested for subsidence under a load of 0.5 kg/cm², it proved to be 0.032.

The coefficient of subsidence of the first monolith, as was noted above, is applied to a layer 1.25 m thick; that of the second—to the next layer of 2.25 m; and that of the third—to the last layer of the compressible zone, 4.0 m thick (see Fig. 4). Hence the total subsidence of the structure, calculated by the formula given above, proved equal to:

\[ S = 0.06 \cdot 125 + 0.037 \cdot 225 + 0.032 \cdot 400 = 28.63\ \text{cm}. \]

If one uses the accepted classification regulating the scope of anti-subsidence measures, this stratum should be assigned to the second category of subsidence. In this case, according to the existing norms and rules for construction on loess-like soils, it will be necessary to apply the appropriate measures ensuring the stability of structures (Building Standards, 1954), or to carry out soil compaction within the limits of the compressible zone.

If, however, the entire thickness of the subsiding loam is included in the calculation and the coefficient of relative subsidence is determined under a load of 3 kg/cm², then for the present example the total amount of subsidence will be

\[ 0.07 \cdot 1500\ \text{cm} = 105\ \text{cm}, \]

that is, 3.5 times greater than the real subsidence of the structure calculated by our method. Accordingly, rather complicated measures will have to be designed in order to preserve the stability of the structures (Building Standards, 1954).

In the Kuznetsk Basin, no cases have yet been observed in which buildings with strip foundations, transmitting to the soil small static loads of up to 2 kg/cm², had settlements exceeding 25–30 cm. This confirms quite well the reality of our method for calculating the probable subsidence of structures erected on soils subsiding under load.

Conclusion

In conclusion, it is appropriate to emphasize the probable limits of applicability of our method and of the method of Yu. M. Abelev.

The method proposed by us, as was stressed above, is applicable for determining the real magnitude of subsidence of structures designed on soils subsiding under load, when the subsidence depends on the additional load transmitted to the soil and extends within the limits of the compressible zone for any duration of soaking of the soil with water. This type of soil includes all the cover subsiding loams investigated by us in a number of districts of Kuzbass. The method of Yu. M. Abelev, recommended by the standards, when applied to these soils gives overestimated values of subsidence. Apparently, the method of Yu. M. Abelev should be applied in cases where structures are designed on soils that subside without additional load, which are widespread in the southern regions of our country.

LITERATURE

Abelev Yu. M.—Practice of Construction on Loess-like Soils. ONTI, 1934.

Abelev Yu. M.—Fundamentals of Design and Construction on Macroporous Soils. Stroyvoenmorizdat, 1948.

Denisov N. Ya.—On the Nature of Subsidence Phenomena in Loess-like Loams. Publishing House “Soviet Science,” 1946.

Denisov N. Ya.—Construction Properties of Loess and Loess-like Loams. State Publishing House of Literature on Construction and Architecture, Moscow, 1953.

Instructions for Determining the Subsidence Properties of Macroporous (Loess-like) Soils. Stroiizdat, 1950.

Building Codes and Regulations, Part II. State Publishing House for Literature on Construction and Architecture, Moscow, 1954.

Nifantov, F. P. — The Geology of Unconsolidated Deposits and the Physical Properties of the Mantle Loess-like Rocks of the Kemerovo District of the Kuzbass. Manuscript in the library of the Tomsk Polytechnic Institute, 1951.

Collection Construction on Loess-like Soils. UkrNITO of Builders, 1939.

Corrections and Noted Misprints

In S. A. Stroitelev’s article “Investigation of the Crystallization of Epsomite and Mirabilite,” for technical reasons braces have been replaced by square brackets, for example: [100] instead of {100}, etc.