Swelling Potential of Compacted Expansive Soils

Swelling Potential of Compacted Expansive Soils

ISSN: 2278-0181

Vol. 2 Issue 3, March – 2013

Swelling Potential of Compacted Expansive Soils

Magdi Zumrawi

Department of Civil Engineering, University of Khartoum Khartoum Sudan

ABSTRACT: Determination of swelling potential is quite important in design of foundations on expansive soils. In this paper an attempt has been made to develop a correlation for Swelling Potential, Sq accounting both soil state and soil type representative parameters. The soil state is reflected by placement conditions fac- tors namely Moisture Content, Dry Density, Void Ratio and Surcharge Pressure whereas the soil type is re- flected by the compositional parameters namely Plasticity Index and Clay Content. The Swelling Potential was measured for compacted samples prepared at different water contents and dry densities. Specimens were allowed to swell under various surcharge loads. Analysis of the experimental results demonstrates very clearly a strong linear relationship of Sq with the Factor of initial soil state, Fi, combination of water content, dry den- sity and void ratio. The coefficients of this linear relationship (i.e. constant and slope) were found to depend on clay content, plasticity index and surcharge pressure. It is shown that predicted values of Swelling Potential using the developed equation agree closely with the experimental results of this study and those reported in the literature.

KEYWORDS: Swelling Potential, Placement Conditions, Initial State Factor

1 INTRODUCTION

The Swelling tendencies of expansive soils are quantified by the swelling potential and/or free swell. Swelling potential or volume change is de- fined as the ratio of increase in thickness to the ini- tial thickness of the soil sample compacted at opti- mum moisture content in a consolidation ring and soaked under a token surcharge of 6.9 k Pa, Seed et al. (1962) so that the sample undergoes free swell.

Holtz and Gibbs (1956) developed a simple test called free swell test for the determination of swell potential of a soil. The test is performed by pouring 10cm3 of soil passing 425µm sieve into a graduated cylinder glass jar of 100ml capacity filled with wa- ter. The swollen volume of the soil is observed after

24 hours. The free swell is expressed as a percentage increase in the volume to the original volume of the soil. Soils having a free swell of 100% or more dam- age lightly loaded structures while those having free swell of less than 50% don't pose serious problems to structures.

* The author E-mail address: magdi.zumrawi@yahoo.com

The swelling potential can be directly determined in the laboratory by performing one-dimensional swell-consolidation test in an oedometer or indirect- ly predicted from classification index properties such as plasticity index, clay content, activity, and shrink- age index which have only an indirect bearing on the degree of swelling, Seed et al. (1962), Rao et al. (2004). Some researchers based on experimental da- ta proposed relationships for swelling potential in- volving both placement conditions and index proper- ties.

The aim of this paper is to predict swelling poten- tial of soil from its index properties such as Atter- berg limits, clay content and placement conditions or initial soil state parameters such as water content, dry density, void ratio and surcharge pressure.

1. PREVIOUS INVESTIGATIONS

   Recently greater attention has been given to em- pirical investigations of the swelling behavior of compacted and natural soils, Seed et al. (1962);

   Chen (1988); Nayak and Christensen (1974); Rao et al. (2004). As a result of these investigations, vari- ous forms of empirical equations have been pro- posed which relate swelling potential to certain physical properties of soils, such as consistency lim- its, clay content, initial moisture content and density.

   Seed et al. (1962) proposed a linear relationship between swell percent, S under a surcharge of 6.9 kPa (1 psi) and plasticity index, PI as:

   ISSN: 2278-0181

   Vol. 2 Issue 3, March – 2013

   S 2.16 103 (PI )2.44

   (1)

   Seed et al. (1962) gave another expression for swell percent, S in terms of activity, A and clay con- tent, C as:

   S 3.6 105(A)2.44 (C)3.44

   (2)

   Chen (1988) proposed a relationship for percent- age swell, S of undisturbed soils in terms of plastici- ty index, PI in the form:

   Figure 2. Volume change versus moisture content for samples compacted at constant dry density, Chen (1975).

   S B e A( PI )

   (3)

   where A and B are constants equal to 0.0838 and 0.2558 respectively. For the development of this cor- relation, the surcharge pressure used was 6.9 kPa. The water content varied between 15% to 20% and dry density between 16 and 17.6 KN/m3.

   Some researchers, Chen (1975) and Kassif et al (1965) based on experimental data proposed rela- tionships for swelling potential and initial soil state parameters such as initial dry density and initial wa- ter content. They found that swelling potential de- pends on both initial water content and dry density. It increases with increasing initial dry density and decreases with increasing initial water content as shown in Figures 1, 2& 3.

   Figure 3. Swell percent versus dry density of samples compact- ed at constant moisture content values, Kassif et al (1965).

   Nayak and Christensen (1974) who studied the swelling behavior of compacted expansive soils, in- dicated that the swell potential can best be related to plasticity index, clay content and water content and gave statistical relationships for swelling potential as:

   S (2.29 102 )(I )1.45 C 6.39

   P wi

   (4)

   Figure 1. The relationship between volume change and initial dry density of expansive soil samples compacted at constant moisture content, Chen (1975).

   where, S: is the percent swell, IP: is plasticity index, C: is the clay content, wi: is the initial water content.

   Brackley (1973) studied the variation of the swell percent with the void ratio for samples compacted at equal moisture content values and observed that the swell percent is inversely proportion with void ratio as shown in Figure 4

   ISSN: 2278-0181

   The initial state factor was empiricVaoll.l2yIsfsouerm3, Mulaarcthe-d2013 on basis of the following reasons:

   • As reported by Chen (1975) and Kassif et al (1965) and shown in Figures 1 and 3 that a linear relationship exists between the swell percent and the initial dry density of the soil samples com- pacted at constant moisture content. Hence, the swell percent, S can be assumed to be directly proportional to the dry density, d

     S d

     ; is

     cons tant

     (7)

     Figure 4. Variation of swell percent with void ratio of samples having equal moisture content values, Brackley (1973).

     The previous empirical equations, (1) to (4) give reasonably good results when applied to the particu- lar soils for which they were developed. Further-

     • The experimental data reported by Chen (1975) and Kassif et al (1965) and shown in Figure 2 and

2. indicated that an inverse linear relationship drawn between the swell percent, S and the mois- ture content, w of the soil samples having the same dry density.

more, they are easy to apply as they relate the swell- ing behavior to simple index properties of soils which can be easily determined in any soil laborato-

S 1

w

; d

is cons tan t

(8)

ry. Consequently, these equations re apparently lack the generality necessary to cover a broad range of soil types.

• Brackley (1973) proposed an inverse linear rela-

tionship between the swell percent, S and the void ratio, e of the samples having the same moisture content as shown in Figure 3.

1. A CONCEPT TO PREDICT SWELLING

   S 1

   e

   ; w is cons tan t

   (9)

   Previous investigations of swelling behavior have revealed that the following factors influence swell-

   Hence from the above equations and using d/w instead of d to make the term dimensionless results in:

   ing potential:

   1. Initial soil state parameters or placement con- ditions.

      S d

      1

      w e

      (10)

   2. Surcharge pressure under which swelling is measured.

   3. Type and amount of clay

   4. Consistency limits

   In order to develop quantitative expression for

   From the above equation (10) it can be concluded that the swell percent, S is directly proportional to the initial state factor, Fi. This relationship can be expressed as:

   swelling potential, the initial soil state parameters such as dry density, water content and void ratio can best be combined in a way reflecting the influence of each of them on the swelling potential. Therefore a

   S Fi

   (11)

   new concept was developed; this is called the initial state factor and can be expressed thus:

2. EXPERIMENTAL WORK

   Fi

   d

   1

   e

   (5)

   The primary objective of this paper is to predict

   the swelling potential of expansive soils using index properties such as water content, dry density, void ratio, clay content and plasticity index. To achieve

   where Fi = the initial state factor; d = the initial

   dry density of soil; w = the density of water; = the initial water content of soil; e = the initial void ratio of soil.

   G

   this objective an experimental testing program was conducted on soil samples collected from different locations of expansive soils in Sudan. Soil samples were compacted with wide range of water content and dry density.

   e s 1

   d

   (6)

   where Gs = the specific gravity of soil.

   Preliminary tests were carried out to determine consistency limits (i.e. liquid limit and plastic limit), clay content and specific gravity all carried out in accordance with BS 1377 (1990).

   The Swell potential was measured in the conven- tional oedometer cell performed on compacted soil samples. The prepared specimens were inundated with water and full swelling attained within 2436 hours. The specimens were permitted to swell under various surcharge pressures of 2.5, 7, 25 and 40 Kpa. Forty eight tests were performed for measuring the swelling potential.

3. RESULTS AND DISCUSSION

   The experimental results of the measured swell percent with the measured index properties, initial water content, dry density, void ratio and their com- bination, initial state factor of the tested samples are given in Table 1. The index properties of the soils and the surcharge pressures under which the swell tests were conducted are presented in Table 2. The measured swell percent of the soils samples were observed to be influenced by the initial dry density and water content as well as the clay content, plastic- ity index and the surcharge load under which the sample was tested.

   1. The Linear Relationship

      To investigate the relationship between the initial state factor, Fi and the swell percent, S the tests re- sults obtained in this study were analysed as given in Table 1. The relationship of the analysed data are shown in Figure 5. The plots in this figure and the values of the correlation coefficient, R2 as listed in Table 2 have clearly shown a linear relation between the swell percent, S and the initial state factor for all

      the data analysed. The straight lines shown in the plots of Figure 5 can be expressed as:

      By substituting the above equations (13) IaSSnNd: 2(21748-)0181

      in the general equation (12) and reVaorlr.a2nIsgsueed3, tMoarcehx–2013

      press swell percent as:

      i

      i

      q

      q

      S 24.5 *(q)0..26 (PI * C)1.26 F 7.1*(q)0.22 *( PI * C)0.78 ] (15)

      where: Fi : is the initial state factor, q: is the surcharge pressure (KPa), PI: is the plasticity index, C: is the clay content.

      Table 1. The measured and predicted swell percent and initial

      s_ta_t_e_d_a_t_a_a_n_a_ly_s_e_d_o_f th_e_f_o_u_r_s_o_i_ls_t_e_s_te_d_. Sample w d e Fi Sm* Sc* Sc / Sm

      % g/cm3 % %

      A1 11.9 1.686 0.69 18.9 13.4 13.7 1.02

      A2 12.0 1.610 0.53 27.3 21.6 20.4 0.94

      A3 15.0 1.609 0.67 15.7 10.1 11.1 1.10

      A4 15.3 1.448 0.48 24.3 19.9 18.0 0.91

      A5 15.5 1.492 0.68 14.9 11.5 10.5 0.91

      A6 16.8 1.500 0.69 13.7 9.4 9.5 1.01

      A7 16.8 1.590 0.83 10.3 7.8 6.8 0.87

      A8 17.4 1.521 0.53 18.9 15.8 13.7 0.87

      A9 19.8 1.673 0.57 14.9 8.7 10.5 1.21

      A10 19.8 1.817 0.65 12.6 8.4 8.7 1.04

      A11 23.0 1.561 0.65 10.8 7.1 7.2 1.02

      A12 23.6 1.472 0.83 7.4 6.0 4.5 0.76

      A1_3 2_8_._7 1_.6_4_8 0_.7_8 6_._7 5_.3 3_._9 0_.7_4_ B1 11.8 1.500 0.77 16.5 8.3 8.2 0.98

      B2 13.3 1.590 0.67 17.9 7.9 8.9 1.13

      B3 13.5 1.521 0.74 15.1 6.4 7.3 1.15

      B4 14.7 1.673 0.58 19.4 10.7 9.8 0.92

      B5 16.2 1.817 0.46 24.4 13.3 12.7 0.96

      B6 17.4 1.561 0.70 12.9 6.6 6.0 0.92

      B7 19.0 1.472 0.80 9.7 3.8 4.2 1.11

      B8 19.0 1.648 0.61 14.2 5.9 6.8 1.16

      B9 21.4 1.721 0.54 14.9 6.4 7.2 1.12

      B10 22.6 1.598 0.66 10.8 4.2 4.8 1.15

      B11 26.3 1.580 0.68 8.9 3.4 3.7 1.12

      B12 26.4 1.489 0.78 7.2 3.2 2.8 0.87

      _B_13 2_8_._9 1_.4_6_0 0_.8_2 6_._2 2_.8 2_._2 0_.8_0_ C1 14.3 1.549 0.77 14.1 29.8 28.7 0.96

      C2 16.9 1.465 0.87 10.0 18.8 18.5 0.99

      C3 20.8 1.358 1.02 6.4 12.6 9.8 0.78

      C4 20.5 1.460 0.88 8.1 14.5 14.1 0.97

      C5 20.7 1.568 0.75 10.1 17.2 19.0 1.10

      C6 23.9 1.370 1.00 5.7 10.6 8.1 0.77

      Sq M ( Fi F0 )

      (12)

      C7 23.8 1.458 0.88 7.0 10.7 11.2 1.05

      C8 29.0 1.371 1.00 4.7 5.8 5.7 0.98

      where: F0 is the value of Fi at zero swell percent,

      M is the gradient of the straight line.

      For the swelling potential data, the relationship of F0 and M with surcharge pressure, clay content and plasticity index was plotted in Figure 6. It can be noted that in this figure, increasing in surcharge load will increase F0 and decrease M values, while in- crease in clay content and plasticity index will in- crease F0 and M values. The equation of the best fit line are expressed thus:

      C9 28.4 1.445 0.90 5.7 5.2 8.0 1.54

      _C_10 3_3_._0 1_.3_7_6 0_.9_9 4_._2 3_.8 4_._4 1_.1_6_ D1 13.6 1.553 0.76 14.9 13.1 12.5 0.73

      D2 14.8 1.520 0.80 12.8 11.8 10.0 0.65

      D3 17.0 1.457 0.88 9.8 8.2 6.3 0.61

      D4 20.7 1.362 1.01 6.5 3.7 2.4 0.56

      D5 20.7 1.453 0.89 7.9 5.4 4.1 0.62

      D6 20.7 1.558 0.76 9.9 6.9 6.6 0.75

      D7 20.5 1.645 0.67 12.1 8.4 9.1 0.84

      D8 23.4 1.369 1.00 5.8 2.8 1.6 0.53

      D9 23.9 1.452 0.89 6.8 3.8 2.8 0.62

      D10 24.5 1.550 0.77 8.3 4.4 4.5 0.83

      0

      0

      F 7.1(q)0.22 *PI *C0.78

      M 24.5(q)0.26 *PI *C1.26

      (13)

      (14)

      D11 29.2 1.352 1.03 4.5 1.1 0.8 0.73

      _D_1_2 2_8_._3 1_.4_5_7 0_.8_8 5_._9 1_.3 1_._7 1_.2_8_

      • Sm: the measured swelling potential value

      • Sc : the calculated swelling potential value.

      Figure 5. The linear relationship between swell percent and ini- tial state factor, Fi for the data analyzed of present tests results.

      Table 2. The tested soils index properties and analysis results.

      perimental data reported by previous inveIsStSiNg:a2t2o78r-s0181 were analysed and drawn in Figure 7.VoAl. s2 Icsasune 3b, eMasreche-n2013

      the trend lines shown in plot indicate that there is a direct linear relationship exists between swell per- cent and the Factor, Fi.

      Sample

      p>_P_I_ _C

      _q –

      G Fi0 Mi R2

      	%	%	KP_a
      Soil A	33	30	7.0	2.65	1.66	0.83	0.948
      Soil B	33	30	25.0	2.65	2.80	0.58	0.951
      Soil C	32	61	2.5	2.74	2.50	2.44	0.951
      Soil D	32	61	40.0	2.74	3.77	1.18	0.954

      	%	%	KP_a
      Soil A	33	30	7.0	2.65	1.66	0.83	0.948
      Soil B	33	30	25.0	2.65	2.80	0.58	0.951
      Soil C	32	61	2.5	2.74	2.50	2.44	0.951
      Soil D	32	61	40.0	2.74	3.77	1.18	0.954

      Figure 7. The relationship between swell percent and initial state factor, Fi for the data reported by previous investigators.

      To check the validity of the present proposed cor- relation, the swell percent values obtained from the proposed equations 15 are compared with the meas- ured swell percent values as given in Figure 8 and also in the Table 1. As can be seen from the tabulat- ed results the ratio between the calculated swelling, using the equation 15 and the measured swelling values are in the range 0.80 1.30. In Figure 8 the points are found to fall close to the line of equality indicating good prediction. This is expected because it is the data used for development of proposed re- gression correlation.

      Figure 6. The variation of F0 and M with clay content, C, plas- ticity index, PI and surcharge pressure, q

   2. Verification With The Reported Data

The validity of the linear relationship and the proposed correlation for Swelling Potential is as- sessed by comparing the results obtained in this in- vestigation with the analysis results of the experi- mental data reported by previous investigators for different soils from literature.

Many investigators have conducted swelling po- tential on compacted expansive soils, Chen (1975); Seed et al.,(1962); Nayak and Christensen (1974). Unfortunately, in most cases, the given data are in- sufficient to apply the proposed equation to their soils. However, the data reported by Chen (1975), Kassif et al (1965) and Brackley (1973) in the litera- ture were selected for comparison.

To evaluate the linear relationship between the in- itial state factor, Fi and the swell percent, S, the ex-

Figure 8. Comparison of measured / predicted Swell Percent for tests results.

The data experimental data reported by Chen (1975), Kassif et al (1965) and Brackley (1973) and shown in Figures 1 to 4 were analysed. The Swelling Potential predicted using the proposed equation 15 is

plotted against the measured Swelling Potential for all the data reported. The plot is shown in Figure 9.

7 ACKNOWLEDGMENTS

ISSN: 2278-0181

Vol. 2 Issue 3, March – 2013

Many points are falling close to the line of equality, some of the points are dispersed away from line of equality. This result indicate that there is a good agreement between the measured and predicted swell percent values and this has proved the validity of the developed equation.

Figure 9. Comparison of measured / predicted Swell Percent for data reported by previous investigators.

6 CONCLUSIONS

Experimental work has been carried out to predict the swelling potential of expansive soils from meas- ured soil properties. Several tests to measure the swell potential and index properties were performed on samples compacted to a wide range of water con- tent and dry density.

The initial state parameters of soil such as water content, dry density and void ratio were combined in

The author would like to thank the technical staff of soil mechanics laboratory of civil engineering de- partment, faculty of engineering, university of Khar- toum for their excellent support during the soil test- ing.

8 REFERENCES

British Standards Institution, BS 1377 (1990). Methods of Test for Soils for Civil Engineering Purposes. BSI, London.

Brackley, I. J. (1973). Swell pressure and free swell in a compacted clay. Proc. 3rd Int. Conf. Expansive Soils, Haifa 1, 169176.

Brackley, I.J.A. (1975). Swell Under Load, Proc. 6th Reg. Conf. for Africa on SM and FE, Curban, S.S., 6570.

Chen, F.H. (1975). Foundations on Expansive Soils. Elsevier Scientific Publishing Co., Amsterdam.

Chen, F.H. (1988). Foundation on expansive soils. Developments in Geotechnical Eng. Elsevier Publg. Co.

Holtz, W. G. and Gibbs, H. J. (1956). Engineering properties of expansive clays. Transactions, ASCE 121, 641-677.

Kassif, G., Komornik, A., Wiseman, G. and Zeitlen,

J. G. (1965). Studies and design criteria for structures on expansive clays. Proc. 1st Res. Conf. on Exp. Soils, Texas.

Mohamed, A. E .M. (1986). Microstructure and swelling characteristics of an untreated and lime- treated compacted black cotton soil. Ph.D thesis. University of Strathclyde, Glasgow.

Nayak, N.V. and Christensen, R.W. (1974). Swelling Characteristics of Compactedexpansive Soils, Clays and Clay Minerals, Vol. 19, No. 4,

a way reflecting the influence of each of them on

pp. 251261.

Ra

. V. and Satyanarayan, B. (1965).

swelling potential. This combination was termed the initial state Factor, Fi. Analysis of the tests results clearly demonstrates a direct linear relationship ex- isting between swell percent and the Factor, Fi.

Based on the linear relationship, strong and relia- ble correlation has been established for swell percent with initial state factor, clay content, plasticity index and surcharge pressure.

Comparison of the measured and predicted swell- ing potential values for tests results and the data re- ported by previous investigators, indicates that there is a good agreement between the measured and pre- dicted swelling values.

The swelling potential and the soil properties seemed to exist in good relationship. Simple regres- sion equation for quick prediction of swell percent from the index properties was developed.

nganatham, B

A rational method of predicting swelling potential for compacted expansive clays. Proc. 6th Inter. Conf. Soil Mechanics Foundation Eng. Vol. 1, pp. 92-96.

Rao, A.S., Phanikumar, B.R. and Sharma, R.S. (2004). Prediction of Swelling Characteristics of Remoulded and Compacted Expansive Soils Using Free Swell Index, Quarterly Journal of Engineering Geology and Hydrogeology, 37: 217226.

Seed, H. B., Woodward, R.J., and Lundgren, R. (1962). Prediction of swelling potential for compacted clays. J. ASCE, Soil Mechanics and Foundation Division, Vol. 88, No. SM-3, Part 1, pp. 53-87.

Zumrawi, M. M. E. (2000). Performance and design of expansive soils as road subgrade. Ph.D. thesis, Changan Univ., China.