Shear strength of compacted Chlef sand: effect of water content, fines content and others parameters

The tests were carried out on the sand of Oued Chlef (Chlef, Algeria), containing 0.5% of non-plastic silt with I_p = 5.02%. The sand was sieved to 0.08mm to obtain the fines, and then cleaned to obtain clean sand. The various parameters such as the maximum void ratio (e_max) (ASTM D4253 [7]) and the minimum void ratio (e_min) (ASTM D4254 [8]). The principal properties and the chemical analysis of the sand, the silt, and the sand-silt mixtures used in this laboratory research work are presented in Tables 1 and 2.

The grain size distribution curve of clean sand, silt and sand-silt mixtures was obtained on the basis of ASTM D422-63 [6] standard are presented in Fig. 1. Various studies were carried out on the sand of Chlef, we cite some work of Arab et al. [4], Belkhatir et al. [10], Belkhatir et al. [11], Benssaleh et al. [13], Della et al. [21], Djafar Henni et al. [22], Hazout et al. [24], Mahmoudi et al. [27]. Clean sand was mixed with 10, 20, 30 and 40% fines to obtain a sand-silt mixtures. Figure 2 shows the photography of the sand, silt and sand-silt mixtures. Figure 3 shows the photography of clean sand and sand-silt mixtures prepared by the wet deposition method.

Figure 1

Figure 2

Figure 3

In order to study the influence of the fines content and the sample preparation method on shear strength, a total of 120 direct shear tests were carried out at three normal stresses 100, 200 and 400 kPa. Sixty (60) tests are performed in a medium dense state (RD = 65%) and sixty (60) in a dense state (RD = 80%). The tests were performed using a square direct shear box 60 x 60 mm². The initial sample height was 25 mm. The test consisted to placing a sample in the shear box and subjected it to a vertical load N that represented the normal stress applied (100, 200 and 400 kPa) and a horizontal load T which was gradually increased. The direct shear test allowed measuring the peak and residual shear strength corresponding to every normal stress. All samples were prepared in three layers by compaction a known mass of the materials studied in the direct shear box to reach target void ratio (already calculated). The samples were prepared at four water contents w = 0, 1, 2 and 3% and at two relative densities (RD = 65% and 80%) under three normal stresses 100, 200 and 400 kPa. Two methods were used to reconstitute the sample, the dry deposition and the wet deposition methods. In the wet deposition method, the dry sand was mixed thoroughly with a water content of w = 1, 2 and 3% until a homogeneous soil sample was obtained. In the dry deposition method, the sand was deposited in the dry state (w = 0%). To achieve the two relative densities, the sample was divided into three layers. Each layer was compacted to achieve the dense state (RD = 80%), however, no compaction was necessary to achieve the medium dense state (RD = 65%), so only the sample surface was levelled off. All the tests were conducted at constant displacement rate of 1.00 mm/min according to ASTM D3080 [9]. The shear stress was recorded as a function of horizontal displacement up to an average shear strain of 7.5 mm.

The effect of fines content on the shear strength of Chlef sand in dense state (RD = 80%) and medium state (RD = 65%) sand was investigated as constituting five sand-silt mixtures F_c = 0, 10, 20, 30 and 40% under three normal stresses 100, 200 and 400 kPa. Figures 4a, 5a, 6a, 7a, 8a and 9a, illustrates the influence of the fines content on the shear strength of sand for w = 0%, it is noted that the maximum resistance decreases with the increase of the fines content these results are in good agreement with the results published in the literature on Chlef sand ([4], [5], [10], [11], [19]). The three samples with 0, 10 and 20% of the fines shows an increase of the shear resistance up to 2 mm and 2.5 mm of deformation then the resistance tends to decrease and stabilizes at 5 mm deformation. While the samples with 30% and 40% of the fines content have a small peak of resistance for deformations greater than 3 mm due to the increase in the amount of fines in the sand. Figures 4b, 5b, 6b, 7b, 8b and 9b show the vertical displacement versus the horizontal displacement. It is very clear that the increase of the relative density decreases the contractancy phase and increases the dilatancy phase whereas the increase of the normal stress increases the contractancy phase and reduces the dilatancy phase. Theses figures show that the increase in the fines content increases the contractancy phase and reduces the dilatancy phase due to the role of the fines content in amplification the contractancy phase and the five samples show a contractancy phase during small deformations (0 to 2 mm) followed by a dilatancy phase for large deformation (2 mm to 7.5 mm).

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figures 10a and 10b illustrate the Mohr-Coulomb failure line that represents the relationship between the maximum shear stress (t_max) and the normal stress (s_n) according to the following expression: (1)tmax=sn tgj+c{\rm{t}}_{{\rm{max}}} = {\rm{s}}_{\rm{n}} \,{\rm{tgj}} + {\rm{c}} Where c and j are the cohesion and the internal friction angle, respectively.

Figure 10

From these figures, it is possible to evaluate the mechanical characteristics c and φ. It is observed for the sand with medium dense state (RD = 65%) and the dense state (RD = 80%) that the slope of the failure line has a greater value 0.89 and 0.97 and this slope decrease with the increase of the fines content. Another observation was observed in this study, the difference between the resistances increases with the increase of the normal stress, that is to say the influence of the fines fraction becomes more pronounced at a great normal stress. Figures 11a and 11b present the mechanical parameters of sand versus the fines content. It is observed from these figures that the increasing of fines content from 10% to 40% leads to increase in the cohesion in logarithm manner from 0 to 15.09 kPa for the medium dense sample and from 0 to 18.86 kPa for the dense sample, and the angle of friction decrease in linear manner from 41.67° to 33.82° for the medium dense sample and from 44.13° to 34.05° for the dense sample, the values of the cohesion and the friction angle are illustrated in Table 3. Our tests result are in good agreement with those found by Flitti et al. [23] where the cohesion of Chlef sand-silt mixtures increases and the friction angle decreases with the increase of the fines content ([4]).

Figure 11

Figures 12a and 13a illustrate the variation of shear stress versus horizontal displacement under shear stress of 200 kPa and two relative densities 65% and 80%. It is very clear that shear strength decreases with the increase the water content from 0 to 3%. For the sample with a water content of 0, 1 and 2% the resistance increases to a value between 2.5 mm and 3.5 mm deformation, then this resistance tends to decrease to a minimum value. For the sample with 3% of water content the behaviour of the dense sand behaves in the same manner as dry loose sand which the mobilized shear stress increases continuously with the increase of horizontal displacement for the three normal stresses. Figures 12b and 13b illustrate the vertical displacement versus the horizontal displacement, these figures show that the increase in the water content has a very clear influence on the volume behaviour, the samples with 0 and 1% of water content show a contractancy phase followed by a dilatancy phase while samples with 2 and 3% water content present only the contractancy phase due to the increase in the water content in the soil.

Figure 12

Figure 13

Figures 14a and 14b illustrate the Mohr-Coulomb failure line that represents the maximum shear stress versus the normal stress. From these figures, it is possible to evaluate the mechanical characteristics c and φ. It is observed that the slope of the failure line is very important for the dry sand and decrease with the increase of the water content and the shear resistance increases with the increase of the normal stress. Figure 15 shows that friction angle decreases linearly from 41.67° to 36.13° for the dense sample with a relative density of 65%, for a relative density of 80% the friction angle decreases from 44.13° to 37.23°, the value of the cohesion and the friction angle are shown in Table 4. The decrease in the friction angle for the two relative densities is due to the presence of water between the grains of the sand and the amount of water creates instability between the grains of the sand.

Figure 14

Figure 15

Figures 16a and 16b show the influence of the water content on the maximum shear strength, the results show that the samples prepared by a relative density RD = 80% gives a greater resistance than of RD = 65% and the maximum shear stress decrease linearly with the increase of water content according to the following expression: (2)tmax=A(w)+B{\rm{t}}_{{\rm{max}}} = {\rm{A}}({\rm{w}}) + {\rm{B}}

Figure 16

The results show that the coefficient A decrease with the increase of the density and the normal stress, and against the coefficient B increase with the increasing of the density and the normal stress. The coefficient A and B are presented in Table 5.

Figures 17a, 17b, 19a and 19b show the difference between the effect of the fines content on shear strength in the wet and dry state. It is observed that the behaviour in the wet state is similar at the behaviour of loose dry sand, the shear strength of fines fraction from 0 to 40% in the dry state is greater than the samples prepared in the wet state with w = 3%, this difference in shear strength due to the increase of the amount of water content in the soil used in this study, in addition the effect of fines content in dry sand and 3% of water content decrease the strength of the sand-silt mixtures. It is observed that increasing the relative density from 65% to 80% increases the shear strength. The effect of adding 3% water content for sand-silt mixtures varies between F_c = 0 to 40% is presented in Figures 18a, 18b, 20a and 20b. It is noted that the mixture of sand and silt prepared in the dry state (w = 0%) has a small contractancy phase followed by a dilatancy phase, while the same samples prepare at w = 3% of water content and with F_c = 30 and 40% of fines content presented as the contractancy phase due to increase the amount of water in the mixtures.

Figure 17

Figure 18

Figure 19

Figure 20

Figures 21a and 21b show the Mohr-Coulomb failure line that represents the maximum shear stress versus the normal stress. It is noted that increasing the water content in sand-silt mixtures decreases the mechanical characteristics c and φ, however the addition of the water content in the clean sand decreases the mechanical characteristic c and φ, 3% of water content influences in a significant manner on the mechanical characteristics of the sand-silt mixtures for a range varies between 0 and 40% of silt.

Figure 21

Figures 22a and 22b illustrate the influence of fines on the maximum shear strength, the results show that the samples prepared by the air pluviation method (AP) gives greater resistance than the samples prepared by the moist tamping method (MT), it is very clear that the maximum resistance decreases in a linear manner with the increase of the fines content according to the following expression: (3)tmax=A(Fc)+B{\rm{t}}_{{\rm{max}}} = {\rm{A}}({\rm{F}}_{\rm{c}} ) + {\rm{B}}

Figure 22

Tables 6 and 7 illustrate the coefficients A, B and the corresponding coefficient of determination (R²) for the Chlef sand-silt mixtures for the moist tamping method and the air pluviation method.

The difference between the method of sample preparation, moist tamping and air pluviation is illustrated in Figures 23a and 23b, it is noted that the shear strength of clean sand increases with increasing density and confining pressure. The air pluviation method gives greater strengths compared to the samples prepared by the method of moist tamping. The samples prepared by the air pluviation method show a peak of resistance between 2 and 3.5 mm of horizontal displacement while the samples prepared by the moist tamping method do not show a peak of resistance and the clean sand behaves the same behaviour of dry loose sand.

Figure 23