Laboratory Investigation on the Strength ... - Springer Link

KSCE Journal of Civil Engineering (2009) 13(1):15-22 DOI 10.1007/s12205-009-0015-x

Geotechnical Engineering

www.springer.com/12205

Laboratory Investigation on the Strength Characteristics of Cement-Sand as Base Material Sungmin Yoon* and Murad Abu-Farsakh** Received April 15, 2008 /Revised September 22, 2008/Accepted November 17, 2008

···································································································································································································································

Abstract Soil cement as pavement base has been used in practice since 1915 due to its effectiveness in preventing pumping of the subgrade soils and its relatively high strength. In this research, the factors affecting the strength of cement-sand were investigated based on laboratory compaction tests and unconfined compression tests. The test results showed that the dry density, moisture content (w), water to cement ratio (w/c), initial void ratio (eini), and cement content (Cw) are the main controlling factors in the behavior of cementsand. A simple empirical correlation method is proposed to estimate the unconfined compressive strength of the cement-sand based on cement content, initial moisture content, and water to cement ratio. The tube suction tests were also performed in an effort to evaluate the water susceptibility of cement-sand, as the strength of cement soil can be affected by the soaking condition at field. The tube suction tests showed that the cement-sand can be classified as good quality base material based on the dielectric value. Keywords: soil cement, unconfined compressive strength, tube suction, pavement base ···································································································································································································································

1. Introduction Cement-stabilized soils have been widely used as base, subbase and/or subgrade materials in pavement structure applications due to the effectiveness in improving their engineering properties as construction materials. Soil cement especially has been used as a beneficial substitute for conventional base materials. The use of soil cement base/subbase can prevent pumping of fine subgrade soils under saturation and heavy traffic load and can provide a strong support for asphalt or concrete pavement structures (ACI Committee 230, 1990). Many researchers have studied the soft clay stabilization using cement, while a mechanism of cement-sand has not yet been well defined (ACI Committee 230, 1990; Mohammad et al., 2000; Chun and Kim, 2001; Guthrie et al., 2002; Kim et al., 2003; Kim et al., 2004; Haeri et al., 2005; Park et al., 2007; Zhang and Tao, 2007). Previous researchers have reported that cement content (Cw), curing time (t), original soil water content (w), total soil water content to cement ratio (w/Cw), and aftercuring void ratio (eot) are the governing factors that describe the engineering behavior of cement-clay (Bergado et al., 1999; Miura et al., 2001; Lorenzo and Bergado 2004). Lee et al. (2001) reported that the behavior of the cement-clay with lower cement content is similar to an over-consolidated soil, and the cementclay with higher cement content behaves like a soft rock. Based on their findings, Lee et al. (2001) proposed a constitutive model for cement-clay, introducing a new parameter (bonding stress

ratio, m) to account for the effect of cementation within the framework of the critical state concept. Abdulla and Kiousis (1997(a) and (b)) also suggested a multi-phase constitutive model to describe the behavior of cement-sand based on micromechanical approach. In their model, the cement-sand is assumed as a composite material consisting of sand, cement, and pore water. The elastoplastic models were developed for each following phases: a) elastoplastic response of the sand, b) elastoplastic response of the cementation before breaking, c) interaction between the phases, and d) breaking of the cement bonds. Then a constitutive model to describe an overall behavior of the cement-sand was developed using assembling the individual phase models. There has been a valid concern of the moisture susceptibility of soil-cement for use as base material. Stavridakis (2005) reported that the soaking condition plays a major role in the strength and the durability of the cemented sand.

2. Objective and Scope The objective of this research is to evaluate the factors affecting the strength and the performance of cement-sand through laboratory unconfined compression tests, as it is the most widely used property of cement-soil in the evaluation of the degree of stabilization. Samples of silty sand mix prepared at four different cement contents were compacted with different moisture contents, and a

*Member, Research Associate, Louisiana Transportation Research Center, Baton Rouge, LA 70808, USA (Corresponding Author, E-mail: [email protected]) **Associate Professor, Research, Louisiana State University, Baton Rouge, LA 70808, USA (E-mail: [email protected]) − 15 −

Sungmin Yoon and Murad Abu-Farsakh

series of unconfined compression tests was conducted to evaluate their strength characteristics. Based on the laboratory tests results, the effects of dry density, molding moisture content, initial void ratio, and cement content of the cement-sand on the unconfined compressive strength were investigated. In addition, a rational correlation method was proposed to estimate the unconfined compressive strength of the cement-sand based on these factors. Furthermore, the water susceptibility of cementsand was evaluated using the tube suction test.

Table 2. General Physical Properties of Cement (provided by manufacturer) Properties

Values

Specific Gravity (Gs)

3.15

Time of set, Initial (minutes)

45

Time of set, Final (minutes)

375

Compressive Strength, 3 days (kPa)

9997

Compressive Strength, 7 days (kPa)

17030

3. Materials and Test Procedures 3.1 Soil The soil investigated in this study is classified as SM (Silty Sand) and A-2 according to the Unified Soil Classification System (USCS) and the AASHTO soil classification, respectively. The general soil properties are summarized in Table 1, and the grain size distribution is shown in Fig. 1. Table 1. General Soil Properties of Tested Sand Soil Properties

Values

Specific Gravity (Gs)

2.68

Uniformity of coefficient (Cu)

6.8

Coefficient of gradation (Cg)

1.7

Soil Classification (USCS) Soil Classification (AASHTO)

3.3 Compaction Tests and Unconfined Compression Tests on Unstabilized Sand and Cement-Sand The cement content ratio can be expressed as follows, weight of cement Cw = ----------------------------------------- × 100 dry weight of soil

(1)

A series of standard Proctor tests were performed on unstabilized sand and cement-sand with different cement contents (Cw = 0, 8, 10, and 12%) in accordance with the ASTM D-698 (ASTM, 2007) test procedure. The samples prepared using standard Proctor tests were extruded from a mold and cured in a humidity room. After seven days of curing, unconfined compression tests were conducted on the cured samples in accordance with ASTM D-1633, method A (ASTM, 2000) test procedure.

Silty Sand, SM

3.4 Tube Suction Test The tube suction test is generally used to evaluate the material moisture susceptibility. For the tube suction tests, three specimens with 12% cement content were compacted at the moisture content of about 8.5% with the compaction energy according to the standard proctor compaction test method (ASTM D 698) inside 152 mm diameter plastic molds. The specimens were 152 mm in diameter and 203 mm in height. Twenty two holes with the diameter of 15 mm were drilled into the bottom of the mold, as shown in Fig. 2. A filter paper was placed on the bottom of the mold. The specimens were dried with the plastic mold in an oven and maintained at 60oC for 48 hours. The specimens were removed from the oven and allowed to cool at room temperature (about 20oC). Six initial dielectric value readings were measured using a surface dielectric probe on the dry specimens. Then the

A-2

Fig. 1. Grain Size Distribution of Tested Sand

3.2 Cement The cement used in this study is Portland cement Type II (moderate sulfate resistance), manufactured by Holcim. The general engineering and physical properties are summarized in Table 2. − 16 −

Fig. 2. Plastic Mold for Tube Suction Test KSCE Journal of Civil Engineering

Laboratory Investigation on the Strength Characteristics of Cement-Sand as Base Material

specimens were place in a flat-bottomed plastic container. The container was filled with water. The water depth was about 6 mm above the bottom of the specimens. The water depth was maintained throughout the tube suction. Dielectric readings were measured daily at six locations on each specimen surface for a total of ten days (See Fig. 3). The highest and lowest readings were disregarded for variability reduction.

based on theoretical formulation of the overall void ratio of mixture consisting of soils with two different grain sizes. According to Lade et al. (1998), when particles with small size are added to a large size particle matrix, the overall void ratio decreases until all the voids within the large particle soils are filled with small particles, which means the dry density increases up to a certain mixing ratio of small particles to large particles. Fig. 5 shows the schematic diagram of theoretical formulation of mixture. Optimum moisture content of the cement-sands shows slightly lower value (OMC = 10.5%) than that of nonstabilized sand (OMC = 11.5%). This may be attributed to lower overall void ratio.

Fig. 3. Tube Suction Test Arrangement

4. Analysis of Test Results and Factors Affecting the Strength of Cement-Soil 4.1 Compaction Tests Dry density of the compacted soil is one of the main factors that influence the strength of the cement-sand. In addition, the water is essential to achieve maximum density and to aid in hydration of the cement. Fig. 4 presents the compaction curves obtained for the nonstabilized sand and cement-sands prepared of different cement contents (Cw = 0, 8, 10, and 12%).

Fig. 5. Schematic Diagram of Theoretical Formulation of Mixture (after Lade et al., 1998)

Fig. 4. Relationship between Dry Density and Moisture Content

As shown in Fig. 4, the dry density (γd) increases with an increase in the cement content (Cw). This can be explained Vol. 13, No. 1 / January 2009

4.2 Effect of Dry Density Cement-sand samples prepared using the standard Proctor compaction tests were cured for seven days in a humidity room, and unconfined compression tests were conducted on these samples. The results of unconfined compression tests are presented in Figs. 6a through 6c. Fig. 7 shows the typical failure mode of the cement-sand by uniaxial unconfined compression. The relationship between the dry density and unconfined compressive strength of tested samples is presented in Fig. 8. Generally, the strength increases with the increase of dry density. However, the highest strength does not occur at the highest dry density. This is due to the factor that the water to cement ratio (w/ c) is one of the major controlling factors affecting the strength of the cement-soil. More detailed discussion will follow in a later section. The aforementioned reason can explain why higher strengths are measured at the dry side of the compaction curve for the samples with the same dry density.

− 17 −


Fig. 6. Stress-Strain Relations from Unconfined Compression Tests: (a) 8% Cement Content, (b) 10% Cement Content, and (c) 12% Cement Content

4.3 Effect of Moisture Content The moisture contents are plotted versus the measured unconfined compression strength, as shown in Fig. 9. As shown in the figure, the unconfined compressive strength has a positive correlation with the moisture content on the dry side and a negative correlation on the wet side. The fact that the highest strength was not measured at the maximum dry density, but instead at the dryer side of the optimum moisture content, is interesting to notice. Detailed discussion will be presented in the following section. Fig. 7. Typical Failure Mode of Cement-Sand

4.4 Water to Cement Ratio (w/c) The engineering properties of cement-soil are usually improved as a result of the hydration process of Portland cement. Water − 18 −

KSCE Journal of Civil Engineering


by weight and the unconfined compressive strength of cementsoil samples. The optimum water to cement ratios that correspond to the highest strength are about 0.85, 1.05, and 1.25 for the sand sample mixed with 12, 10, and 8% cement contents, respectively. The figure also shows that the optimum water to cement ratio for samples with higher cement content is lower than those with lower cement content. This can be explained as when more cements are added to the cement-soil mixture, the cement to water contact area will increase, i.e., more waters in the cement-soil will be used in cement hydration process.

Fig. 8. Relationship between Dry Density and Unconfined Compressive Strength

Fig. 10. Relationship between Water to Cement Ratio by Weight and Unconfined Compressive Strength

Fig. 9. Relationship between Moisture Content and Unconfined Compressive Strength

is essential for cement hydration; Therefore, water in the cementsoil should be enough to ensure the complete hydration. In typical concrete design, the minimum water to cement ratio by weight ((w/c)min) is 0.42 (Mindess et al., 2002). This value can not be used directly in cement-soil, as the cement particles represent a small portion in cement-soil matrix and are not completely in contact with water. However, if there is too much water, it will lead to greater voids and hence lower strength. Therefore, the strength of cement-soil depends on the combination of water to cement ratio and compacted dry density. Fig. 10 shows the relationship between water to cement ratio Vol. 13, No. 1 / January 2009

4.5 Effect of Cement Ratio (Cw) Fig. 11 presents the relationship between the cement content and the unconfined compressive strength of cement-sand samples. It should be noted here that the compressive strengths were estimated by linear interpolation from different molding moisture contents. As shown in the figure, the compressive strength has a positive correlation with the cement content. However, the effect of higher cement is not distinct in the sample with high moisture content. This may be due to lower compacted dry density that will result in greater voids and hence lower strength. 4.6 Effect of Initial Void Ratio (eini) Since the dry density can represent the degree of compactness for soils with same grain size distribution, the void ratio may be a more proper factor for use in evaluating both the degree of compactness and the frictions between the soil particles. The initial void ratios of the samples were calculated using the wet and dry densities and the specific gravity. Fig. 12 depicts the general trend of unconfined compressive strength with respect to the initial void ratio for the cement-sand samples. The figure

− 19 −


Fig. 11. Relationship between Cement Content by Weight and Unconfined Compressive Strength

factors. Both dry density (γd) and initial void ratio (eini) present the sample compactness, however they are interrelated factors with the relationship of γd = ( Gs γw ) ⁄ ( 1 + e ) , and the void ratio can represent the soil compactness, regardless of soil specific gravity (Gs). Therefore, the soil compactness is considered in the proposed correlation using an initial void ratio (eini). The moisture content (MC) also interrelates with the void ratio and the water to cement ratio (w/c). It is of more importance to the compressive strength in providing enough water for cement hydration. Therefore, the water to cement ratio (w/c) is used to account for the effect of water in cement-sand. Based on the laboratory unconfined compression test results, a possible correlation between these factors combined, as ( Cw ⁄ ( eini ( w ⁄ c ) ) ) ratio, and the UCS was investigated, as shown in Fig. 13. The figure shows that there is a linear correlation between the combinations of these factors and the unconfined compressive strength. A simple regression analysis conducted on the collected data yielded the following correlation model to estimate UCS of cement-sand: Cw qu = 0.62 × Pa -------------------eini ( w ⁄ c )

(2)

where Pa = reference pressure (atmospheric pressure).

Fig. 12. Relationship between Initial Void Ratio and Unconfined Compressive Strength

clearly indicates that the unconfined compressive strength increases with the decrease of the sample initial void ratio. 4.7 Correlation Model to Estimate Unconfined Compressive Strength of Cement-Sand As discussed earlier in previous sections, the major influencing factors that affect the compressive strength of cement-sand are: material compactness (γd, eini), moisture content (MC), water to cement ratio (w/c), and cement content (Cw). Three factors (eini, w/c and Cw) out of five factors (γd, MC, eini, w/c and Cw) investigated in this study were used to propose an empirical correlation between the compressive strength and the different

Fig. 13. Relationship between Cw/eini (w/c) and Unconfined Compressive Strength

4.8 Tube Suction Tests The measured dielectric value (DV) from tube suction tests for three cement-sand samples with cement content of 12% and initial water content of about 8.5% are plotted versus time, as shown in Fig. 14. As seen in the figure, the DV values were stabilized after seven days of soaking, and the final DV values

− 20 −



are all less than 5.5. According to the recommendation reported by Barbu and Scullion (2006), as shown in Table 3, the cementsand with cement content of 12% can be classified as a good quality base.

and lower moisture content, as compared to those with optimum moisture contents and maximum dry densities. This may be due to lower w/c ratios in these samples. 3. Based on the results of the unconfined compression tests, the following factors were noticed as the major factors influencing the unconfined compressive strength of the cement-sand: a) material compactness, represented by void ratio (eini), b) water to cement ratio (w/c), and c) cement content (Cw). An empirical correlation between these factors combined, as ( Cw ⁄ ( eini ( w ⁄ c ) ) ) ratio, and the UCS was proposed. 4. The results of the tube suction tests performed on cementsand with 12% cement confirmed that the moisture susceptibility of the material is low. Therefore, the material can be classified as good quality base, according to recommendations suggested by Barbu and Scullion (2006).

Acknowledgements This research project is supported by the Louisiana Transportation Research Center and Louisiana Department of Transportation and Development. The authors gratefully acknowledge the valuable help of Zhongjie Zhang and Mark Morvant, of LTRC.

References Fig. 14. Variation of Dielectric Values with Time Table 3. DV Criteria for Base Material (Barbu and Scullion, 2006) Final DV < 10 10 – 16 > 16

Classification Good quality bases Marginal quality Poor quality

5. Conclusions The effects of moisture content, dry density (γd), w/c ratio, and cement content (Cw) on the strength of cement-sand were investigated using a series of unconfined compression tests. The cement-sand specimens were prepared using standard Proctor compaction tests and were tested after seven days of curing. Based on the test results, the following conclusions are made: 1. The standard Proctor maximum dry density (γd) of the cement-sand increases with the increase of cement content (Cw). This is because the finer cement particles will fill the voids in sands that have larger particle sizes. This will take place until all voids in sands are filled with cement particles. In this research, the cement content (Cw) between 0 and 12% and dry density (γd) shows a positive relation. 2. The specimens with higher dry densities (γd) generally show higher strength. However, the highest strength was observed at the dry side of the compaction curve; i.e., the highest strength was observed on the sample with lower dry density Vol. 13, No. 1 / January 2009

Abdulla, A. and Kiousis, P. (1997a). “Behavior of cemented sands – I. Testing.” International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 21, pp. 533-547. Abdulla, A. and Kiousis, P. (1997b). “Behavior of cemented sands – II. Modeling.” International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 21, pp. 549-568. ACI Committee 230 (1990). “State-of-the-art report on soil cement.” ACI Materials Journal, American Concrete Institute, Vol. 87, No. 4, pp. 395-417. American Society for Testing and Materials. (2000). ASTM D 1633. Standard Test Methods for Compressive Strength of Molded SoilCement Cylinders, Philadelphia, PA. American Society for Testing and Materials. (2007). ASTM D 698. Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort (12,400 ft-lbf/ft3 (600 kN-m/m3)), Philadelphia, PA. Barbu, B. and Scullion, T. (2006). Repeatability and Reproducibility Study for Tube Suction Test. FHWA/TX-06/5-4114-01-1, Texas Transportation Institute. Bergado, D.T., Ruenkrairergsa, T., Taesiri, Y., and Balasubramaniam, A.S. (1999). “Deep soil mixing to reduce embankment settlement.” Ground Improvement Journal, Vol. 3, No. 3, pp. 145-162. Chun, B.S. and Kim, J.C. (2001). “A study of soil cement properties by using soilcrete stabilizer.” Journal of Korean Geo-Environmental Society, Vol. 2 No. 4, pp. 73-81. Guthrie, W., Sebesta, S., and Scullion, T. (2002). Selecting Optimum Cement Contents for Stabilizing Aggregate Base Materials, FHWA/ TX-05/7-4920-2, Texas Transportation Institute. Haeri, S., Hosseini, S., Toll, D., and Yasrebi, S. (2005). “The behavior of an artificially cemented sandy gravel.” Geotechnical and Geological Engineering, Vol. 23, pp. 537-560.

− 21 −


Kim, B.I., Wee, S.H., Lee, S.H., and Kim, Y.U. (2003). “Strength characteristics of soil-cement mixed with inorganic solidification liquid.” Journal of the KSCE, KSCE, Vol. 23, Issue 3C, pp. 135141. Kim, J.H., Won, J.Y. Kim, S.P., and Chang, P.W. (2004). “Phase changes of soil-cement mixture using fall cone and heat of hydration.” Journal of the Korean Geotechnical Society, Vol. 20, Issue 9, pp. 2532. Lade, P.V., Liggio, C.D., and Yamamuro, J.A. (1998). “Effects of nonplastic fines on minimum and maximum void ratios of sand.” Geotechnical Testing Journal, ASTM, Vol. 21, No. 4, pp. 336-347. Lee, S., Lee, K.H., Yi, C.T., and Jung, D.S. (2001). “A constitutive model for cemented clay in a critical state framework.” Journal of the Korean Geotechnical Society, Vol. 17, Issue 1, pp. 119-129. Lorenzo, G. and Bergado, D. (2004). “Fundamental parameters of cement-admixed clay—new approach.” Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Volume 130, Issue 10, pp. 1042-1050. Mindess, S., Young, J., and Darwin, D. (2002). Concrete, 2 ed., Prentice

Hall. Miura, N., Horpibulsok, S., and Nagaraj, T. (2001). “Engineering behavior of cement stabilized clay at high water content.” Soils and Foundations, Vol. 41, No. 5, pp. 33-45. Mohammad, L., Raghavandra, A., and Huang, B. (2000). “Laboratory performance evaluation of cement-stabilized soil base mixtures.” Transportation Research Record, No. 1721, TRB, National Research Council, Washington, D.C., pp. 19-28. Park, S.S., Kim, Y.S., and Lee, J.C. (2007). “Unconfined compressive strength of fiber-reinforced cemented sands by fiber reinforcement form.” Journal of the Korean Geotechnical Society, Vol. 23, Issue 8, pp. 159-169. Stavridakis, E. (2005). “Evaluation of engineering and cement-stabilization parameters of clayey-sand mixtures under soaked conditions.” Geotechnical and Geological Engineering, Vol. 23, pp. 635-655. Zhang, Z. and Tao, M. (2007). “Build reliable cement-treated subgrade layer.” The Proceedings of 7th International Conference of Chinese Transportation Professionals, May 21-22, 2007, Shanghai, China, pp. 79-90.

− 22 −
