fbpx

Reference Library

Soilworks products are the industry’s top standard due to our insistence on creating high performance soil stabilization and dust control products that stand up to rigorous testing – both in the lab and in the field. Our commitment to quality and performance has led to our involvement and testing in hundreds of real-world situations. The following library of reports, presentations, specifications, approvals and other similar documents provide you, our customer, the transparency and dependable assurance that is expected from Soilworks.

ASCE – Characterization of Cement-Polymer-Treated Soils under Repeated Loading (TPD2401004)

Characterization of Cement-Polymer-Treated Soils under Repeated Loading

Prince Kumar, A.M.ASCE1; Anand J. Puppala, Ph.D., P.E., D.GE, F.ASCE2; Surya Sarat Chandra Congress, Ph.D., A.M.ASCE3; and Jeb S. Tingle, P.E., M.ASCE4

 

Abstract: The mechanistic-empirical pavement design guide (M-EPDG) recommends the use of resilient modulus (MR) for characterization of subgrade soils. Subgrade soils may not always have enough strength and stiffness to support the pavement structure. Therefore, a certain type of soil improvement method using cement, lime, or other stabilization techniques is often needed to enhance the strength and stiffness properties of weak subgrade soils. The cement-stabilized soils show brittle behavior under compression loading, which can induce cracking in overlying pavement layers. In general, polymer-treated soils show a semiductile or ductile behavior. It is important to look for combined cement and polymer treatments to address brittle behavior issues as well as moisture susceptibility while maintaining strength and moduli properties. A research study was conducted to understand the strength, resilient, and ductile behaviors of sandy soils treated with cement and a combination of cement and vinyl acetate ethylene (VAE) copolymer. Engineering tests such as unconfined compressive strength (UCS) and resilient modulus tests were conducted on both control and treated soil specimens cured for 7 days. Tests were conducted on specimens before and after immersing in water bath for 4 h to investigate the moisture susceptibility. In these tests, an increase in UCS was observed after cement and cement-VAE treatments as compared to control soil specimens. Results showed that cement-VAE-treated soils exhibited an increase in the axial strain at failure, indicating the semiductile behavior compared to cement-treated specimens. An improvement in the resilient moduli was observed after treatments. Subsequently, two of three-parameter models were used to analyze resilient modulus formulations with stress conditions and determined the regression constants. In conclusion, the study revealed that the use of VAE copolymer improved the stress-strain responses of cement-treated soils and imparted closer to the semiductile behavior, which will reduce cracking in overlying pavement structures. DOI: 10.1061/JMCEE7.MTENG-16394. © 2024 American Society of Civil Engineers.

Author keywords: Subgrade; Sand; Soil stabilization; Cement; Polymer; Resilient modulus; Unconfined compressive strength (UCS).

Introduction and Background

Natural subgrade soils do not always possess adequate strength and stiffness properties to support the pavement structure subjected to repetitive traffic loading. One of the infrastructure solutions is to replace local soils with better quality borrowed soils to improve the strength and stiffness of the subgrade layer; however, this alterna- tive depends on several factors such as the transportation costs and the availability of quality borrow material near the construction site (Abu-Farsakh et al. 2015). Hence, other alternatives including dif- ferent ground improvement methods are often explored to stabilize weak subgrades to improve their strength and stiffness properties. Calcium-based stabilizers are widely used in practice to improve

 

1Ph.D. Candidate, Zachry Dept. of Civil and Environmental Engineering, Texas A&M Univ., College Station, TX 77843. ORCID: https://orcid

.org/0000-0001-9352-2905. Email: prince.kumar@tamu.edu

2A.P. and Florence Wiley Chair Professor, Zachry Dept. of Civil and Environmental Engineering, Texas A&M Univ., College Station, TX 77843 (corresponding author). ORCID: https://orcid.org/0000-0003-0435-6285. Email: anandp@tamu.edu

3Assistant Professor, Dept. of Civil and Environmental Engineering, Michigan State Univ., East Lansing, MI 48824. ORCID: https://orcid

.org/0000-0001-5921-9582. Email: surya@msu.edu

4Senior Research Civil Engineer and Program Manager, Geotechnical and Structures Laboratory, US Army Engineer Research and Development Center, Vicksburg, MS 39180. Email: Jeb.S.Tingle@erdc.dren.mil

Note. This manuscript was submitted on February 1, 2023; approved on September 13, 2023; published online on January 26, 2024. Discussion period open until June 26, 2024; separate discussions must be submitted for individual papers. This paper is part of the Journal of Materials in Civil Engineering, © ASCE, ISSN 0899-1561.

the strength and stiffness of subgrade soils (Petry and Little 2002; Little and Nair 2009; Abu-Farsakh et al. 2015; Puppala 2016, 2021; Behnood 2018). Soil stabilization mechanisms depend on the sta- bilizer used and soil type in the field conditions. Cement stabilization mechanisms include cation exchange, flocculation and agglomeration, cementitious hydration, and pozzolanic reaction proc- esses whereas lime stabilization mechanisms have all the aforemen- tioned processes except cementitious hydration reactions (Prusinski and Bhattacharja 1999; Little and Nair 2009; Puppala 2016).

Cement- or lime-stabilized soil specimens show brittle behavior under compression and flexural loading from traffic loads, which can induce rutting and various types of pavement cracking (Puppala et al. 2005; Little et al. 2009; Onyejekwe and Ghataora 2015). Mitigating reflective cracking on pavement surfaces caused by the brittle behavior of cement-stabilized soil is always a major challenge (Rezaeimalek et al. 2018). Therefore, stabilization methods that can induce ductility behavior and increase the toughness of treated soils are often sought. Synthetic polymer treatment is one such method that can be used as a soil-curing agent with cement to reduce the brittle behavior of cement-treated materials (Tingle et al. 2007; Onyejekwe and Ghataora 2015; Zhang et al. 2022). Available literature reveals that polymer stabilization could also potentially increase strength and stiffness, reduce permeability, and increase the durability properties of soils (Santoni et al. 2002; Iyengar et al. 2012; Kolay et al. 2016; Georgees et al. 2017; Huang et al. 2021; Kumar et al. 2022a, b). Zhang et al. (2022) used ethylene-vinyl copolymer (EVA) as a curing agent with cement additive for loess soil stabilization and ob- served significant improvements in the unconfined compressive strength properties of loess soil. Similarly, Onyejekwe and Ghataora (2015) reported an increase in UCS after polymer and

 

cement treatment of lime query fines. The EVA-treated soil specimens showed ductile failure mode. The strain at failure increased after adding synthetic polymer to cement-treated soil samples. Scanning electron microscopy or SEM studies showed that a large amount of calcium-silicate-hydrate (CSH) gel formed after the application of a mixture of EVA and cement stabilizers compared to cement treatment alone (Zhang et al. 2022).

Yaowarat et al. (2022) studied the effect of cement and polyvinyl alcohol (PVA) treatment on the mechanical properties of recycled concrete aggregates (RCA) and found that the UCS of cement-treated RCA material reduced after the addition of 0.5% and 1% of PVA but increased in case of 1.5% and 2% of PVA when compared to the cement-treated RCA. The resilient moduli of cementtreated RCA samples were higher than PVA-stabilized samples (Yaowarat et al. 2022). This was due to the cement treatment improved strength and stiffness of the material, therefore, rigidity increased. Conversely, the PVA improved the shear strength and toughness properties of treated soils. Based on a comprehensive literature review, the authors noted that no study has been conducted on understanding the resilient behavior of cement-VAE-treated soils using repeated load triaxial (RLT) testing. Hence, it is essential to study resilient moduli characteristics of present cement-VAE-treated soils to assess the improvements emanating from polymer treatments.

The design and performance of the pavement structures depend on the stiffness of the subgrade soil. The resilient modulus is a measure of stiffness for subgrade soils and is considered as a key design parameter for flexible pavement systems under repetitive traffic loads (AASHTO 2004; Papagiannakis and Masad 2008; Puppala 2008). American Association of State Highway and Transportation Officials (AASHTO)’s M-EPDG recommends the use of the resilient modulus for characterizing subgrade soils (AASHTO 2004; Puppala 2008). The resilient modulus is defined as the ratio of cyclic axial stress (σcyclic) to resilient axial strain (εr) as presented mathematically in Eq. (1) (Seed et al. 1962)

conducted on soil specimens before immersing them in the water bath and after immersing them in the water bath to investigate the moisture susceptibility of engineering properties. Two three-parameter MR characterization models were then used to analyze and model the resilient modulus test results, and regression constants were determined by conducting regression analysis. The effect of confining pressure, deviator stress, stabilizer, and stabilizer dosages on the resilient moduli of soils were discussed in detail.

 

Experimental Program

 

Materials

This study considered two sandy soils obtained from Bryan, Texas. The present research study followed US Department of Defense standard unified facilities criteria (UFC) 3-250-11 (UFC 2004, 2020) for soil stabilization with cement and polymer. Basic geo-technical tests such as wet sieve analysis, hydrometer analysis, Atterberg limits, and specific gravity tests were conducted as per respective ASTM standards (ASTM 2019). Fig. 1 shows particle size distribution curves for both soils. Soil-1 and soil-2 were classified as poorly graded sand (SP) and silty sand (SM), respectively as per the unified soil classification system (USCS). The maximum dry unit weight (MDUW) and optimum moisture content (OMC) of untreated and treated soil mixtures were determined using a modified Proctor compaction test (ASTM 2019). Table 1 presents the physical properties of both soils.

Commercially available cement type-I and vinyl acetate-ethylene copolymer were used as stabilizers for the present laboratory studies. This VAE copolymer used in the current research is a commercially available synthetic polymer and is in liquid form, has milky white color with mild odor, and is 100% dispersible. It has 55% solid content and 45% water.

 

M

σcyclic

R ¼  ϵr

ð1Þ

 

Several test methods were developed to determine the resilient modulus of unbound materials in the laboratory (Puppala 2008). However, the repeated load triaxial test is regularly used to measure the resilient modulus of unbound materials in the laboratory (Puppala 2008; Han and Vanapalli 2016). This test was designed to simulate the stress induced by traffic loading through a series of cyclic deviator stresses applied to the specimen at different confining pressures (Puppala 2008; AASHTO 2017). Past researchers observed that the resilient modulus of unbound and bound materials depends on several variables including deviator stress, confining pressure, dry density, moisture content, soil properties, stabilizer type, stabilizer dosage, curing condition, and curing time (Seed et al. 1962; Fredlund et al. 1977; Puppala et al. 1996, 2003; Drumm et al. 1997; Puppala 2008; Solanki et al. 2010; Abu-Farsakh et al. 2015; Kumar et al. 2013). Several M characterization models such as two- and three-parameter models were developed by past researchers and these models were used to characterize the resilient behavior of untreated and treated subgrade soils (Uzan 1985; Pezo 1993; Puppala et al. 2003; Puppala 2008Solanki et al. 2010; Abu-Farsakh et al. 2015; Kumar et al. 2022a). The present research was designed and focused on assessing the strength, resilient, and ductile behaviors of cement-VAE-treated sandy soils. For comparisons, both untreated and cement-treated sandy soils were also considered in this study. Extensive experimental- tal tests including modified Proctor compaction tests, unconfined compressive strength tests, and repeated load triaxial tests were conducted on both untreated and treated soils. The UCS and RLT tests were conducted at different cement and VAE dosages. Tests were

conducted on soil specimens before immersing them in the water bath and after immersing them in the water bath to investigate the moisture susceptibility of engineering properties. Two three- parameter MR characterization models were then used to analyze and model the resilient modulus test results, and regression con- stants were determined by conducting regression analysis. The ef- fect of confining pressure, deviator stress, stabilizer, and stabilizer dosages on the resilient moduli of soils were discussed in detail.

 

Experimental Program

 

Materials

This study considered two sandy soils obtained from Bryan, Texas. The present research study followed US Department of Defense standard unified facilities criteria (UFC) 3-250-11 (UFC 2004, 2020) for soil stabilization with cement and polymer. Basic geo- technical tests such as wet sieve analysis, hydrometer analysis, Atterberg limits, and specific gravity tests were conducted as per respective ASTM standards (ASTM 2019). Fig. 1 shows particle size distribution curves for both soils. Soil-1 and soil-2 were clas- sified as poorly graded sand (SP) and silty sand (SM), respectively as per the unified soil classification system (USCS). The maximum dry unit weight (MDUW) and optimum moisture content (OMC) of untreated and treated soil mixtures were determined using a modi- fied Proctor compaction test (ASTM 2019). Table 1 presents the physical properties of both soils.

Commercially available cement type-I and vinyl acetate ethyl- ene copolymer were used as stabilizers for the present laboratory studies. This VAE copolymer used in the current research is a com- mercially available synthetic polymer and is in liquid form, has milky white color with mild odor, and is 100% dispersible. It has 55% solid content and 45% water.

 

Fig. 1. Grain size distribution of soils used in this study.

 

                                                                                                             

 

 

 

Table 1. Properties of soils used in this study

 

 

Parameters                                                       Soil-1                                Soil-2

Sand (%)                                                          96.3                                   82.4

Silt (%)                                                               3.7                                   16.2

Clay (%)                                                             0.0                                      1.4

% finer than #200 sieve                                  3.7                                   17.6

USCS classification                                          SP                                      SM

Specific gravity, Gs                                           2.67                                   2.67

Plasticity index (%)                                          NP                                      NP Note: NP = nonplastic.

 

Specimen Preparation Protocols

Both control and treated soil specimens were prepared according to the US Department of Defense UFC 3-250-11 (UFC 2004, 2020) standard and AASTHO T 307-99 (AASHTO 2017) standard. Dry soil was mixed homogeneously with water for untreated/control specimens. The predetermined amount of cement was homogeneously mixed with dry soil until a uniform color was obtained and then precalculated water was thoroughly mixed with the soil-cement mixture until a uniform product was obtained for cement-treated soil specimens. For cement-VAE-treated soil specimens, the precalculated amount of VAE was then thoroughly mixed with the soil-cement-water mixture until a uniform product was achieved. The amount of water already present in the VAE was accounted for while preparing the treated soil mixtures. Each mixture was then statically compacted into a cylindrical mold having a height of 142 mm and a diameter of 71 mm at respective optimum moisture contents and maximum dry unit weights. All specimens were com- pacted within half an hour after the soil mix was prepared to prevent the initial set of soil mixtures. A sample calculation for the mass of soil, water, cement, and VAE is provided in Table 2 for cement-VAE-treated SP soil. Similar procedures and calculations were adopted for other soil specimen groups.

±

Treated soil specimens were prepared at different dosages of cement and VAE by weight of dry soil and these specimens were cured in an air-drying room (22°C ± 2°C and 50% relative humid- ity) for 7 days. Additionally, cement-treated soil specimens were prepared and cured in a moist room (22°C 2°C and 100% relative humidity) for 7 days. Table 3 summarizes stabilizer dosages and test notations followed for each soil mixture. Triplicate specimens of control and treated soils were prepared for each mixture for conducting UCS and MR tests. Overall, four sets of specimen groups were prepared, the first and second sets were prepared for dry and wet conditions for UCS tests, respectively; and the third and fourth sets were prepared for dry and wet conditions for RLT tests, respectively. Two types of control conditions have been considered for the untreated soil: one set immediately after compaction (control-1)

 

Table 2. Specimen preparation calculation for cement-VAE-treated SP soil Parameters

Table 3. Stabilizer dosages and notations used in this study

 

and the other set prepared at the same curing conditions as stabilized soil (control-2) for ensuring similar curing conditions for untreated and treated soils in this research study.
The curing condition at a temperature of 22°C 􀀁 2°C and 50% relative humidity could be considered as an air-dry process and this has been recommended over the moist curing method for polymer-treated soils (Santoni et al. 2002). Also, this air-dry curing process in the laboratory closely simulates actual field conditions during military construction procedures (Tingle and Santoni 2003; Newman and Tingle 2004; Santoni et al. 2005). Under this process, the evapora-tion of water from soil specimens hardens the polymer-soil matrix.
Unconfined Compressive Strength and Repeated Load Triaxial Tests
The UCS tests have been commonly used to evaluate the effectiveness of chemical stabilizers for road construction. The set of three specimens were cured and tested according to the dry test procedure. The remaining three specimens were immersed in the water bath for 4 h and then UCS tests were conducted as per ASTM D1633. The UCS tests were performed with a constant strain rate of 0.90% /min. The RLT tests were performed by following the AASHTO T 307-99 standard method for determining the resilient modulus of both untreated and treated soils (AASHTO 2017). All the tests were conducted using the repeated load triaxial test equipment shown in Fig. 2. These tests were conducted at three confining pressures (41.4 kPa, 27.6 kPa, and 13.8 kPa) and five deviatoric stresses (13.8 kPa, 27.6 kPa, 41.4 kPa, 55.2 kPa, and 68.9 kPa) to simulate different stress conditions induced due to traffic loading. First, the specimen was preconditioned by applying 500 load cycles with a cyclic stress of 24.8 kPa at confining pressure of 41.4 kPa. Second, the specimen was tested for 15 loading sequences that includes 100 cycles of load at different combinations of deviator stress and confining pressure. The details of the testing sequence are provided in Table 4. The preconditioning sequence helps in reducing initial irregularities between the top platen and the test specimen. Also, it helps in preventing the effects of the time interval between com-paction and loading (AASHTO 2017). The cyclic deviator stress was then applied as a haversine-shaped loading with a 0.1 s load duration, followed by a 0.9 s relaxation period, as shown in Fig. 3. The load was measured using a submersible load cell. Axial deformation of the soil specimen under repeated loading were measured using two linear variable differential transducers (LVDTs). The LVDTs were attached to the rods at 180° diametrically opposite to each other on top of the triaxial cell. Average of recoverable de-formation obtained from both LVDTs was then used in the calculation of resilient modulus.

 

Fig. 2. Repeated load triaxial test apparatus.

 

Table 4. Testing sequence for subgrade soil

 

Fig. 3. Haversine-shape and duration of each repeated load cycle.

 

using two linear variable differential transducers (LVDTs). The LVDTs were attached to the rods at 180° diametrically opposite to each other on top of the triaxial cell. Average of recoverable de-formation obtained from both LVDTs was then used in the calculation of resilient modulus.

Test Results Analysis and Discussion

Compaction Test Results
The modified Proctor compaction test results of untreated and treated SP and SM soils are shown in Figs. 4 and 5, respectively. Maximum dry unit weights and the optimum moisture contents for both untreated and treated soils are summarized in Table 5. An increase in the dry unit weight of SP soil was observed after cement treatment. Only a minor increase in the maximum dry unit weight of SM soil was observed after cement stabilization. The reason for this increase might be due to the higher specific gravity of cement relative to soils. Another reason for the increase could be the size of cement particles. Cement consists of very fine-grained particles

 

Fig. 4. Compaction curves for untreated and treated SP soil. ZAV = zero air void; and S = degree of saturation.

Fig. 5. Compaction curves for untreated and treated SM soil. ZAV = zero air void; and S = degree of saturation.

Table 5. Maximum dry unit weights and optimum moisture contents for untreated and treated soils

 

which fill the voids between sand particles which increases the dry unit weight of sand mixtures (Al-Aghbari et al. 2009). The addition of VAE to the soil-cement mixture further increased the dry unit weight of both SP and SM soils after treatment. It can be argued that the VAE copolymer provided a lubrication effect on soil particles which increased the dry unit weight of soils. Both cement and cement-VAE treatments decreased the OMC of both SP and SM soils; however, the decrease in OMC was more after the cement-VAE treatment compared to the cement treatment. The degree of saturation corresponding to OMC varied between 50% to 57% for SP soil and 62% to 65% for SM soil.
Unconfined Compressive Strength (UCS) Test Results
Fig. 6 presents the UCS test results of control and treated SP and SM soils. Tests were conducted before and after immersion of the soil specimen in the water bath. The UCS of the SP and SM control-1 soil specimens were observed to be 0.01 and 0.04 MPa, respectively. It can be observed that the UCS of untreated SP and SM soils (control-2) increased after 7 days of air-dry curing when compared to control-1 soil. This increase was because of enhancement in strength due to evaporation of moisture in the air-dry curing condition. The UCS of control-2 SM soil was higher than that of control-2 SP soil and authors attribute to higher percent finer particles in SM soil which contributed to higher matric suction during air-dry curing process. Previously, researchers studied the concept of soil suction and observed that soil matric suction increases with a reduction in water content (Fredlund and Rahardjo 1993). This concept of soil suction was used to explain results. Both control-1 and control-2 soil specimens disintegrated immediately after immersing them in water.
This study also conducted UCS tests on moist cured cement-treated specimens as per the ASTM D1632 method and these results are used here for comparison purposes. It was observed that moist-cured cement-treated specimens showed higher UCS values than air-dry cured specimens. This variation was attributed to the unavailability of water for the hydration reactions in the air-dry cured cement-treated soil specimens. UCS test results showed that cement-treated SP and SM soil specimens observed higher UCS values when compared to control soils. The UCS of cement-treated soils reduced after immersion in the water bath and the reduction in UCS values was more in the case of air-dry cured cement-treated specimens than moist cured cement-treated specimens.
UCS test results indicated a significant increase in UCS value when VAE copolymer was combined with cement treatment.

Fig. 6. Unconfined compressive strength (UCS) results for untreated and treated soils before and after immersion in water bath (a) SP soil; and (b) SM soil.

Fig. 7. Stress-strain curves for untreated and treated (a) SP; and (b) SM soils before immersion in water bath.

 

This increase in UCS value can be attributed to two factors: an increase in strength due to semi-dry conditions from the evaporation of moisture in the air-dry curing room and cementitious reactions or/and hardening of the additives within the soil matrix. The UCS values of SP+C3+P5-AD and SP+C5+P5-AD were observed to be 3.70 MPa and 4.59 MPa, respectively, and these were reduced to 0.33 and 1.02 MPa, respectively, after water bath immersion. It is important to note that the percent reduction in UCS values was more in the case of cement-VAE-treated soil specimens as compared to cement-treated soil specimens. However, the UCS values of SP+C3+P5-AD and SP+C5+P5-AD were higher than SP-C3-AD and SP-C5-AD, respectively, after water bath immersion. The probable reason for the higher loss of UCS in the case of cement-VAE treatment compared to cement treatment can be explained. Cement performed better in binding soil particles compared to VAE copolymer during moisture exposure, which resulted in this observation. In both cases, strength loss was also observed due to a change in state from semi-dry to moist after water bath immersion. SM soil also showed similar UCS trends as those of SP soil and these results can be seen in Fig. 6(b).
Figs. 7(a and b) show stress-strain curves obtained from UCS tests for untreated and treated SP and SM soil specimens, respectively. In these figures, specimens tested in dry condition (prior to immersion in the water bath) are presented. Table 6 presents a summary of axial strains at the failure of untreated and treated SP and SM soils. The axial strains at the failure have substantially increased after adding VAE to cement-treatments of SP and SM soil specimens. This indicates an increase in the ductility behavior of cement-treated soil specimens after mixing with VAE copolymer. It can be inferred that the addition of VAE to cement treatment has resulted in the semiductility behavior during axial loading when compared to cement-treated soils. This positive outcome may potentially reduce the shrinkage cracking from stiffer cement-treated soil layers on the pavement surfaces in the real field conditions as cement-treated soils often experience and induce brittle cracking in the field conditions.
Repeated Load Triaxial (RLT) Test Results
The RLT tests were conducted on both untreated and treated soil specimens before and after immersing them in the water bath. The RLT tests were conducted on triplicate soil specimens and average values were used in the plots and discussions as follows. Figs. 8 and 9 present stress and strain response of untreated SP soil under repeated load, respectively. It can be observed that the cyclic load was applied in Haversine load form for 0.1 s. The axial strain response was used to determine resilient strain, which was used in the calculation of resilient modulus. Typical variations of resilient moduli of untreated and treated SP and SM soils with various stress conditions are presented in Figs. 10 and 11, respectively. It can be observed that both control and treated SP and SM specimens showed an increase in resilient moduli with an increase in confining pressure. Control-1 SP and SM soil specimens showed an increase in resilient moduli with an increase in deviator stress. These results indicate the stress-hardening behavior of these soils when specimens are subjected to cyclic deviator stress loading. Under the stress-hardening phenomenon, the material becomes stronger with an increase in deviator stress. The resilient moduli values of control-2

 

Table 6. Strain at failure for untreated and treated soils

Fig. 8. Deviator stress response of untreated SP soil for the last 5 cycles of sequence number 5 of AASHTO T 307-99.

Fig. 9. Axial strain response of untreated SP soil for the last 5 cycles of sequence number 5 of AASHTO T 307-99.

 

SP and SM soil specimens were higher than control-1 specimens of same soils, which can be attributed to the semi-drying conditions of control-2 specimens resulting in higher moduli.
Cement-treated SP and SM soils showed stress-hardening behavior at both 3% and 5% cement dosages in both curing conditions, both before and after immersions in the water bath. An increase in resilient modulus was also observed with an increase in cement dosage. The resilient moduli of cement-treated soil specimens were reduced after immersing in the water bath at both dosages, and the reduction in the case of 3% cement treated soil was more than those at 5% cement content. This is expected because higher cement dosage results in stronger specimens with higher cementation bonding between soil particles. A significant reduction in resilient moduli of cement-treated soils cured in air-dry conditions was observed after immersion in water. Conversely, a slight reduction in the resilient moduli of cement-treated soils was observed when cured in a moist curing room. This observation is due to the lack of water for facilitating chemical reactions in air-dry curing conditions when compared to moist curing conditions. Also, loss of soil matric suction induced during air-dry curing contributed to this moduli variation.
The resilient moduli of cement-VAE-treated SP and SM soil specimens showed stress-hardening behavior at both cement dosages, before and after immersing them in the water bath. As expected, an increase in resilient modulus was observed with an increase in cement dosages for both soils. A substantial reduction in resilient moduli was observed after soaking these specimens in water and this reduction was expected and is attributed to loss of soil matric suction during saturation under immersion.
Figs. 12 and 13 present a comparison between the resilient moduli of untreated and treated SP and SM soil specimens, respectively, at confining pressure of 13.8 kPa and deviator stress of 41.4 kPa (i.e., 13th sequence in AASHTO T 307-99 test method). This stress condition was considered as it was recommended by previous studies to determine design resilient modulus value for flexible pavements (Jones and Witczak 1977; Solanki et al. 2010). It can be observed that the control-1 soil specimen showed the lowest resilient modulus for both soils. Resilient modulus of cement-VAE-treated SP soil was observed to be higher than control soils before immersion in the water bath. The SP soil treated with a mixture of 5% cement and 5% VAE showed the highest resilient modulus. Results showed that the control-2 SM soil specimen showed the highest resilient modulus and remained almost same after treatment with 5% cement and 5% VAE before immersion in the water bath. Control soil specimens disintegrated completely after immersion in the water bath, therefore, 100% reduction was observed in resilient modulus of both control soils. A substantial reduction was observed after immersing cement-VAE-treated soil specimens under water. As expected, this percent reduction was lower at higher cement

Fig. 10. Resilient modulus of untreated and treated SP soil before and after immersion in water bath (a) control-1; (b) control-2; (c) 3% cement;(d) 5% cement; (e) 3% cement þ5% polymer; and (f) 5% cement þ5% polymer.

Fig. 11. Resilient modulus of untreated and treated SM soil before and after immersion in water bath (a) control-1; (b) control-2; (c) 3% cement;(d) 5% cement; (e) 3% cement þ5% polymer; and (f) 5% cement þ5% polymer.

Fig. 12. Resilient modulus of 13th sequence for untreated and treated SP soil.

Fig. 13. Resilient modulus of 13th sequence for untreated and treated SM soil.

 

dosage treatment. The retained resilient modulus of cement-VAE-treated soils after moisture conditioning is still high indicating the need for chemical treatment to mitigate moisture susceptibility of soils. Overall, it can be mentioned that the cement-VAE treatment has not significantly resulted in the resilient property improvements of the present sandy soils.

Modeling of Resilient Moduli of Treated Soils
Different types of MR characterization models such as two-parameter and three-parameter models were developed for characterizing unbound materials (Seed et al. 1962; Uzan 1985; Pezo 1993; Mohammad et al. 1999; Ooi et al. 2004; Puppala 2008). The two-parameter models have some limitations and are highlighted by several researchers (e.g., Uzan 1985; Mohammad et al. 1999; Puppala 2008). To overcome the limitations of the two-parameter models, three-parameter models were developed. Although these models were developed for characterizing unbound materials, several researchers used them for characterizing stabilized materials (Solanki et al. 2010; Abu-Farsakh et al. 2015; Bhuvaneshwari et al. 2019; Jose et al. 2021; Kumar et al. 2022a). The current research study considered two three-parameter models developed by Pezo (1993) and National Cooperative Highway Research Program (NCHRP) project 1-28A (AASHTO 2004), mathematical equations are provided in Eqs. (2) and (4), respectively. These models were used for characterizing MR test results of both untreated and treated soils. Pezo (1993) considered both confining pressure and deviator stress in the model. The NCHRP project 1-28A developed model (also called the three-parameter universal model or M-EPDG model) accounts for bulk stress and octahedral shear stress. The M-EPDG recommends the three-parameter universal model for characterizing unbound materials

fx

where, MR = resilient modulus; k1p, k2p and k3p = regression con-stants for Pezo (1993) model; k1m, k2m and k3m = regression constants for M-EPDG model; σd = deviator stress = σ1 − σ3; θ = bulk stress = σ1 þ σ2 þ σ3; Pa = atmospheric pressure; and τoct = octahedral shear stress.

fx

 

All three model constants were determined by conducting multiple-linear regression analyses on the resilient moduli test results. Tables 7 and 8 present regression constants and the coefficient of determination (R2) for both untreated and treated SP and SM soils, respectively. This research study obtained a minimum value of R2 of 0.94 for SP soil and 0.92 for SM soil, which indicates an excellent fit between the model predictions and experimental test results.
Based on the results presented in Tables 7 and 8, the following observations are noted:

  • The model constant, k1p, increased after cement-VAE treatment (in dry test condition) as compared to control-2 specimen for SP soil. On the contrary, this constant was decreased after cement-VAE treatment compared to control-2 specimen for SM soil. It was observed that the value of k1p reduced after immersing in the water bath for both soils. Similar trends as k1p were observed for k1m for both soils.
  • The coefficient, k2p, which describes the stiffening behavior of the treated material with an increase in confining pressure, was less than 1.0 for all untreated and treated SP and SM soils, and these values varied between 0.125 and 0.531. This indicates that the effect of confining pressure decreases with an increase in resilient modulus.
  • The value of k2p decreased after air-drying control soil specimens (control-2) as compared to control-1 specimens for both soils. The cement-VAE treatment (in dry test condition) further reduced the value of k2p as compared to control-2 specimens. This observation shows that the effect of confining pressure on MR decreased after air-drying or treatment. It was observed that the value of k2p increased after immersing cement-VAE-treated soil specimens in the water bath compared to that of before immersion for both soils. Both characterization models showed similar behavior of the k2 parameter for all soil groups.

Table 7. Three-parameter model constants for control and treated SP soils

Table 8. Three-parameter model constants for control and treated SM soils

 

Fig. 14. Measured MR versus predicted MR we (a) Pezo model (Pezo 1993); and (b) M-EPDG model.

 

  • In the Pezo (1993) model, a positive and negative value of the parameter, k3p, represents stress-hardening and stress-softening behavior, respectively. The value of k3p was positive for both soils and it varied between 0.016 and 0.234, which shows stress-hardening behavior. Control-1 specimens of both SP and SM soils showed negative values of k3m for the M-EPDG model. It was observed that the M-EPDG model showed both positive and negative values of k3m for treated SP and SM soil specimens.
  • The coefficient k3p decreased after cement-VAE treatment (in dry test condition) compared to control-2 specimen for both soils and further reduced after immersing in the water bath. It was observed that the value of k3m decreased after 3% cement and 5%VAE treatment and increased after 5% cement and 5% VAE treatment as compared to control-2 specimens for both soils, and these values reduced after immersing in the water bath.
    The predicted resilient modulus values were compared with the measured resilient modulus values, as shown in Figs. 14 and 15 for both SP and SM soils, respectively. Both Pezo (1993) and M-EPDG models reasonably characterized and predicted the resilient modulus values in the back-calculation analysis.

Conclusions and Recommendations
A research study presented in this paper has comprehensively assessed the effectiveness of the mixture of cement and VAE in enhancing the unconfined compressive strength and resilient moduli of sandy soils while reducing the brittle behavior that is usually associated with cement-treated soils. Various variables including cement dosages and curing conditions were studied and addressed. Both UCS and RLT tests were conducted on untreated and treated soils compacted at optimum conditions from modified Proctor test results. The resilient modulus test results were analyzed using three-parameter models such as Pezo (1993) and M-EPDG models. The following conclusions are drawn based on the experimental study results:

  1. Cement and cement-VAE treatments of SP and SM soils have increased the maximum dry unit weight or MDUW compared to untreated soils. The increase in MDUW was more in the case of SP soil as compared to SM soil after cement and cement-VAE stabilizations. The optimum moisture content was reduced slightly after cement and cement-VAE treatments for both soils.
  2. Both cement and cement-VAE treatments improved the unconfined compressive strength properties of both SP and SM soils. The UCS results of cement-treated specimens cured in a moist room were higher than those soils cured in the air-drying room. This was due to the lack of sufficient water for cement hydration reaction when specimens were cured in the air-drying room. An increase in UCS was, however, observed with an increase in cement dosages.
  3. A significant increase in the UCS values of SP and SM soils was reported after cement-VAE treatment. Also, the UCS value was increased with an increase in cement percentage in the cement-VAE mixture. This increase in strength was provided due to a combination of the development of soil matric suction induced due to the evaporation of moisture in soil specimens and the bonding between soil particles created by the mixture of cement and VAE. The UCS values were reduced after immersing treated soil specimens under water. The addition of VAE into cement increased the axial strain at the failure which shows an improvement in ductile behavior after the combined cement and polymer stabilization method as compared to cement treatment alone.
  4. The resilient moduli values of control-2 SP and SM soil specimens were higher than control-1 specimens of same soil, which can be attributed to the semi-drying conditions of control-2 specimens resulting in higher moduli. Control soil specimens disintegrated immediately after immersion in the water bath. Resilient modulus of cement-VAE-treated SP soil was observed to be higher than control soils before immersion in the water bath. The control-2 SM soil specimen showed the highest resilient modulus and remained almost the same after treatment with 5% cement and 5% VAE. A substantial reduction was observed after immersing these cement-VAE-treated soil specimens under water. The retained resilient modulus of cement-VAE-treated soils after moisture conditioning is still high when compared to control-2 soils indicating the role of chemical treatment to mitigate moisture susceptibility of these soils.
  5. Both three-parameter models provided a higher coefficient of determination and obtained an excellent fit with experimental data. Therefore, it is recommended to use both three-parameter models for the determination of the resilient modulus of soils considered in this research study.

The model parameters, k1p and k1m, of cement-VAE-treated SP and SM soils were higher and lower than the same control-2 soil specimens, respectively. The coefficients k2p and k2m decreased after cement-VAE treatment compared to the control specimens for both soils. This shows that the effect of confining pressure or bulk stress on the resilient modulus is reduced after treatment. Both cement-VAE-treated soils showed lower values of k3p coefficient compared to control-2 specimen. No definite trend was observed in the case of k3m coefficient. These k3p and k3m values reduced after immersing in the water bath for both soils.
The present research study observed an improvement of UCS and resilient moduli of SP and SM soils after the treatment with a mixture of cement and VAE and moduli improvements are moderate with the addition of VAE. Furthermore, the VAE additive enhanced the failure behavior of cement-treated specimens from brittle to semi-ductile. Therefore, the mixture of cement and polymer can be recommended for the stabilization of soils considered in this study. Both Pezo (1993) and M-EPDG models precisely predicted the resilient modulus values and the authors recommend these models for soils with characteristics similar to the soils used in the current study.

Data Availability Statement

All data generated or used during the study appear in the form of figures and tables in the published article. Some or all data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments
The authors would like to acknowledge US Army Engineer Research and Development Center, Vicksburg, Mississippi for granting the funds for Research project #W912HZ 20P0090. Also, the authors would also like to acknowledge Mr. Zoheb Faisal, a graduate student for his help during experimental works.

References
AASHTO. 2004. Mechanistic-empirical pavement design guide. Washington, DC: AASHTO.

AASHTO. 2017. Standard method of test for determining the resilient modulus of soils and aggregate materials. AASHTO T 307-99. Washington, DC: AASHTO.

Abu-Farsakh, M., S. Dhakal, and Q. Chen. 2015. “Laboratory characterization of cementitiously treated/stabilized very weak subgrade soil under cyclic loading.” Soils Found. 55 (3): 504–516. https://doi.org/10.1016/j .sandf.2015.04.003.

Al-Aghbari, M. Y., Y. E. A. Mohamedzein, and R. Taha. 2009. “Stabilisation of desert sands using cement and cement dust.” Proc. Inst. Civ. Eng. Ground Improv. 162 (3): 145–151. https://doi.org/10.1680/grim .2009.162.3.145.

ASTM. 2019. Annual book of ASTM standards. West Conshohocken, PA:ASTM.

Behnood, A. 2018. “Soil and clay stabilization with calcium- and non-calcium-based additives: A state-of-the-art review of challenges, approaches and techniques.” Transp. Geotech. 17 (Dec): 14–32. https://doi.org/10.1016/j.trgeo.2018.08.002.

Bhuvaneshwari, S., R. G. Robinson, and S. R. Gandhi. 2019. “Resilient modulus of lime treated expansive soil.” Geotech. Geol. Eng. 37 (1): 305–315. https://doi.org/10.1007/s10706-018-0610-z.

Drumm, E. C., J. S. Reeves, M. R. Madgett, and W. D. Trolinger. 1997. “Subgrade resilient modulus correction for saturation effects.” J. Geotech. Geoenviron. Eng. 123 (7): 663–670. https://doi.org/10.1061/(ASCE) 1090-0241(1997)123:7(663).

Fredlund, D. G., A. T. Bergan, and P. K. Wong. 1977. “Relation between resilient modulus and stress research conditions for cohesive subgrade soils.” Transp. Res. Rec. 642 (Mar): 73–81.

Fredlund, D. G., and H. Rahardjo. 1993. Soil mechanics for unsaturated soils. New York: Wiley.

Georgees, R. N., R. A. Hassan, R. P. Evans, and P. Jegatheesan. 2017. “Resilient response characterization of pavement foundation materials using a polyacrylamide-based stabilizer.” J. Mater. Civ. Eng. 30 (1): 04017252. https://doi.org/10.1061/(ASCE)MT.1943-5533.0002109.

Han, Z., and S. K. Vanapalli. 2016. “State-of-the-art: Prediction of resilient modulus of unsaturated subgrade soils.” Int. J. Geomech. 16 (4):04015104. https://doi.org/10.1061/(ASCE)GM.1943-5622.0000631.

Huang, J., R. B. Kogbara, N. Hariharan, E. A. Masad, and D. N. Little. 2021. “A state-of-the-art review of polymers used in soil stabi-lization.” Constr. Build. Mater. 305 (Oct): 124685. https://doi.org/10 .1016/j.conbuildmat.2021.124685.

Iyengar, S. R., E. Masad, A. K. Rodriguez, H. S. Bazzi, D. Little, and H. J. M. Hanley. 2012. “Pavement subgrade stabilization using poly-mers: Characterization and performance.” J. Mater. Civ. Eng. 25 (4): 472–483. https://doi.org/10.1061/(ASCE)MT.1943-5533.0000612.

Jones, M. P., and M. W. Witczak. 1977. “Subgrade modulus of the San Diego test road.” Transp. Res. Rec. 641: 1–6.

Jose, A., J. M. Krishnan, and R. G. Robinson. 2021. “Resilient and permanent deformation response of cement-stabilized pond ash.” J. Mater. Civ. Eng. 34 (1): 1–14. https://doi.org/10.1061/(ASCE)MT.1943-5533 .0004044.

Kolay, P. K., B. Dhakal, S. Kumar, and V. K. Puri. 2016. “Effect of liquid acrylic polymer on geotechnical properties of fine-grained soils.” Int. J. Geosynth. Ground Eng. 2 (4): 1–9. https://doi.org/10.1007/s40891-016-0071-5.

Kumar, P., A. J. Puppala, S. S. C. Congress, K. Ramineni, and J. S. Tingle. 2013. “Influence of compaction characteristics and moisture exposure on resilient moduli of cement-treated soil.” In Geo-Congress 2023: Soil Improvement, Geoenvironmental, and Sustainability, Geotechnical Special Publication 339, edited by E. Rathje, B. M. Montoya, and M. H. Wayne, 370–380. Reston, VA: ASCE.

Kumar, P., A. J. Puppala, J. S. Tingle, S. Chakraborty, S. Sarat, and C. Congress. 2022a. “Resilient characteristics of polymer emulsion-treated sandy soil.” Transp. Res. Rec. 2676 (9): 1–13. https://doi.org/10.1177 /03611981221087238.

Kumar, P., J. S. Tingle, A. J. Puppala, S. Chakraborty, and S. S. C. Congress. 2022b. “Permeability characteristics of polymer emulsion-stabilized soils.” In Vol. 331 of Geo-Congress 2022, Geotechnical Special Publication 331, edited by A. Lemnitzer and A. W. Stuedlein, 261–270. Reston, VA: ASCE.

Little, D., and S. Nair. 2009. “Recommended practice for stabilization of subgrade soils and base materials.” In National cooperative highway research program. Washington, DC: Transportation Research Board.

Little, D. N., S. Nair, and B. Herbert. 2009. “Addressing sulfate-induced heave in lime treated soils.” J. Geotech. Geoenviron. Eng. 136 (1): 110–118. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000185.

Mohammad, L. N., B. Huang, A. J. Puppala, and A. Allen. 1999. “Regression model for resilient modulus of subgrade soils.” Transp. Res. Rec. 1687 (1): 47–54. https://doi.org/10.3141/1687-06.

Newman, K., and J. S. Tingle. 2004. “Emulsion polymers for soil stabilization.” In Proc., 2004 FAA Worldwide Airport Technology Transfer Conf. Washington, DC: Federal Aviation Administration.

Onyejekwe, S., and G. S. Ghataora. 2015. “Stabilization of quarry fines using a polymeric additive and portland cement.” J. Mater. Civ. Eng. 28 (1): 04015070. https://doi.org/10.1061/(ASCE)MT.1943-5533 .0001324.

Ooi, P. S. K., A. R. Archilla, and K. G. Sandefur. 2004. “Resilient modulus models for compacted cohesive soils.” Transp. Res. Rec. 1874 (1): 115–124. https://doi.org/10.3141/1874-13.

Papagiannakis, A. T., and E. A. Masad. 2008. Pavement design and materials. New York: Wiley.

Petry, T. M., and D. N. Little. 2002. “Review of stabilization of clays and expansive soils in pavements and lightly loaded structures—History, practice, and future.” J. Mater. Civ. Eng. 14 (6): 447–460. https://doi.org /10.1061/(ASCE)0899-1561(2002)14:6(447).

Pezo, R. 1993. “A general method of reporting resilient modulus tests of soils: A pavement engineer’s point of view.” In Proc., 72nd Annual Meeting of the Transportation Research Board. Washington, DC: Transportation Research Board.

Prusinski, J. R., and S. Bhattacharja. 1999. “Effectiveness of Portland cement and lime in stabilizing clay soils.” Transp. Res. Rec. 1652 (1): 215–227. https://doi.org/10.3141/1652-28.

Puppala, A., L. Mohammad, and A. Allen. 1996. “Engineering behavior of lime-treated Louisiana subgrade soil.” Transp. Res. Rec. 1546 (1): 24–31. https://doi.org/10.3141/1546-03.

Puppala, A. J. 2008. Estimating stiffness of subgrade and unbound materials for pavement design. Washington, DC: Transportation Research Board.

Puppala, A. J. 2016. “Advances in ground modification with chemical additives: From theory to practice.” Transp. Geotech. 9 (Dec): 123–138. https://doi.org/10.1016/j.trgeo.2016.08.004.

Puppala, A. J. 2021. “Performance evaluation of infrastructure on problematic expansive soils: Characterization challenges, innovative stabilization designs, and monitoring methods.” J. Geotech. Geoenviron. Eng. 147 (8): 04021053. https://doi.org/10.1061/(ASCE)GT.1943-5606.0002518.

Puppala, A. J., N. Intharasombat, and R. K. Vempati. 2005. “Experimental studies on ettringite-induced heaving in soils.” J. Geotech. Geoenviron. Eng. 131 (3): 325–337. https://doi.org/10.1061/(ASCE)1090-0241 (2005)131:3(325).

Puppala, A. J., A. M. Ramakrishna, and L. R. Hoyos. 2003. “Resilient moduli of treated clays from repeated load triaxial test.” Transp. Res. Rec. 1821 (May): 68–74. https://doi.org/10.3141/1821-08.

Rezaeimalek, S., R. Nasouri, J. Huang, and S. Bin-Shafique. 2018. “Curing method and mix design evaluation of a styrene-acrylic based liquid polymer for sand and clay stabilization.” J. Mater. Civ. Eng. 30 (9): 04018200. https://doi.org/10.1061/(ASCE)MT.1943-5533.0002396.

Santoni, R. L., J. S. Tingle, and M. Nieves. 2005. “Accelerated strength improvement of silty sand with nontraditional additives.” Transp. Res. Rec. 1936 (1): 34–42. https://doi.org/10.1177/0361198105193600105.

Santoni, R. L., J. S. Tingle, and S. L. Webster. 2002. “Stabilization of silty sand with nontraditional additives.” Transp. Res. Rec. 1787 (1): 61–70. https://doi.org/10.3141/1787-07.

Seed, H., C. Chan, and C. E. Lee. 1962. “Resilience characteristics of subgrade soils and their relation to fatigue failures in asphalt pavements.” In Proc., Int. Conf. on the Structural Design of Asphalt Pavements, 611–636. Washington, DC: Transportation Research Board.

Solanki, P., M. M. Zaman, and J. Dean. 2010. “Resilient modulus of clay subgrades stabilized with lime, class c fly ash, and cement kiln dust for pavement design.” Transp. Res. Rec. 2186 (1): 101–110. https://doi.org /10.3141/2186-11.

Tingle, J. S., K. Newman, S. L. Larson, C. A. Weiss, and J. F. Rushing. 2007. “Stabilization mechanisms of non-traditional additives.” Transp. Res. Rec. 1989 (1): 59–67. https://doi.org/10.3141/1989-49.

Tingle, J. S., and R. L. Santoni. 2003. “Stabilization of clay soils with non-traditional additives.” Transp. Res. Rec. 1819 (1): 72–84. https://doi.org /10.3141/1819b-10.

UFC (Unified Facilities Criteria). 2004. Soil stabilization for pavements. UFC 3-250-11. Washington, DC: Dept. of Defense.

UFC (Unified Facilities Criteria). 2020. Soil stabilization and modification for pavements. UFC 3-250-11. Washington, DC: Dept. of Defense.

Uzan, J. 1985. “Characterization of granular material.” Transp. Res. Rec. 1022 (1): 52–59.

Yaowarat, T., S. Horpibulsuk, A. Arulrajah, F. Maghool, M. Mirzababaei, A. S. A. Rashid, and A. Chinkulkijniwat. 2022. “Cement stabilisation of recycled concrete aggregate modified with polyvinyl alcohol.” Int. J. Pavement Eng. 23 (2): 349–357. https://doi.org/10.1080/10298436 .2020.1746311.

Zhang,B.,W.Jiang,Q.Xu,D.Yuan,J.Shan,andR.Lu.2022. “Experimental feasibility study of ethylene-vinyl acetate copolymer (EVA) as cement stabilized soil curing agent.” Road Mater. Pavement Des. 23 (3): 617–638. https://doi.org/10.1080/14680629.2020 .1834442.

Introduction and Background

Natural subgrade soils do not always possess adequate strength and stiffness properties to support the pavement structure subjected to repetitive traffic loading. One of the infrastructure solutions is to replace local soils with better quality borrowed soils to improve the strength and stiffness of the subgrade layer; however, this alterna- tive depends on several factors such as the transportation costs and the availability of quality borrow material near the construction site (Abu-Farsakh et al. 2015). Hence, other alternatives including dif- ferent ground improvement methods are often explored to stabilize weak subgrades to improve their strength and stiffness properties. Calcium-based stabilizers are widely used in practice to improve

 

1Ph.D. Candidate, Zachry Dept. of Civil and Environmental Engineering, Texas A&M Univ., College Station, TX 77843. ORCID: https://orcid

.org/0000-0001-9352-2905. Email: prince.kumar@tamu.edu

2A.P. and Florence Wiley Chair Professor, Zachry Dept. of Civil and Environmental Engineering, Texas A&M Univ., College Station, TX 77843 (corresponding author). ORCID: https://orcid.org/0000-0003-0435-6285. Email: anandp@tamu.edu

3Assistant Professor, Dept. of Civil and Environmental Engineering, Michigan State Univ., East Lansing, MI 48824. ORCID: https://orcid

.org/0000-0001-5921-9582. Email: surya@msu.edu

4Senior Research Civil Engineer and Program Manager, Geotechnical and Structures Laboratory, US Army Engineer Research and Development Center, Vicksburg, MS 39180. Email: Jeb.S.Tingle@erdc.dren.mil

Note. This manuscript was submitted on February 1, 2023; approved on September 13, 2023; published online on January 26, 2024. Discussion period open until June 26, 2024; separate discussions must be submitted for individual papers. This paper is part of the Journal of Materials in Civil Engineering, © ASCE, ISSN 0899-1561.

the strength and stiffness of subgrade soils (Petry and Little 2002; Little and Nair 2009; Abu-Farsakh et al. 2015; Puppala 2016, 2021; Behnood 2018). Soil stabilization mechanisms depend on the sta- bilizer used and soil type in the field conditions. Cement stabilization mechanisms include cation exchange, flocculation and agglomeration, cementitious hydration, and pozzolanic reaction proc- esses whereas lime stabilization mechanisms have all the aforemen- tioned processes except cementitious hydration reactions (Prusinski and Bhattacharja 1999; Little and Nair 2009; Puppala 2016).

Cement- or lime-stabilized soil specimens show brittle behavior under compression and flexural loading from traffic loads, which can induce rutting and various types of pavement cracking (Puppala et al. 2005; Little et al. 2009; Onyejekwe and Ghataora 2015). Mitigating reflective cracking on pavement surfaces caused by the brittle behavior of cement-stabilized soil is always a major challenge (Rezaeimalek et al. 2018). Therefore, stabilization methods that can induce ductility behavior and increase the toughness of treated soils are often sought. Synthetic polymer treatment is one such method that can be used as a soil-curing agent with cement to reduce the brittle behavior of cement-treated materials (Tingle et al. 2007; Onyejekwe and Ghataora 2015; Zhang et al. 2022). Available literature reveals that polymer stabilization could also potentially increase strength and stiffness, reduce permeability, and increase the durability properties of soils (Santoni et al. 2002; Iyengar et al. 2012; Kolay et al. 2016; Georgees et al. 2017; Huang et al. 2021; Kumar et al. 2022a, b). Zhang et al. (2022) used ethylene-vinyl copolymer (EVA) as a curing agent with cement additive for loess soil stabilization and ob- served significant improvements in the unconfined compressive strength properties of loess soil. Similarly, Onyejekwe and Ghataora (2015) reported an increase in UCS after polymer and

 

cement treatment of lime query fines. The EVA-treated soil specimens showed ductile failure mode. The strain at failure increased after adding synthetic polymer to cement-treated soil samples. Scanning electron microscopy or SEM studies showed that a large amount of calcium-silicate-hydrate (CSH) gel formed after the application of a mixture of EVA and cement stabilizers compared to cement treatment alone (Zhang et al. 2022).

Yaowarat et al. (2022) studied the effect of cement and polyvinyl alcohol (PVA) treatment on the mechanical properties of recycled concrete aggregates (RCA) and found that the UCS of cement-treated RCA material reduced after the addition of 0.5% and 1% of PVA but increased in case of 1.5% and 2% of PVA when compared to the cement-treated RCA. The resilient moduli of cementtreated RCA samples were higher than PVA-stabilized samples (Yaowarat et al. 2022). This was due to the cement treatment improved strength and stiffness of the material, therefore, rigidity increased. Conversely, the PVA improved the shear strength and toughness properties of treated soils. Based on a comprehensive literature review, the authors noted that no study has been conducted on understanding the resilient behavior of cement-VAE-treated soils using repeated load triaxial (RLT) testing. Hence, it is essential to study resilient moduli characteristics of present cement-VAE-treated soils to assess the improvements emanating from polymer treatments.

The design and performance of the pavement structures depend on the stiffness of the subgrade soil. The resilient modulus is a measure of stiffness for subgrade soils and is considered as a key design parameter for flexible pavement systems under repetitive traffic loads (AASHTO 2004; Papagiannakis and Masad 2008; Puppala 2008). American Association of State Highway and Transportation Officials (AASHTO)’s M-EPDG recommends the use of the resilient modulus for characterizing subgrade soils (AASHTO 2004; Puppala 2008). The resilient modulus is defined as the ratio of cyclic axial stress (σcyclic) to resilient axial strain (εr) as presented mathematically in Eq. (1) (Seed et al. 1962)

conducted on soil specimens before immersing them in the water bath and after immersing them in the water bath to investigate the moisture susceptibility of engineering properties. Two three-parameter MR characterization models were then used to analyze and model the resilient modulus test results, and regression constants were determined by conducting regression analysis. The effect of confining pressure, deviator stress, stabilizer, and stabilizer dosages on the resilient moduli of soils were discussed in detail.

 

Experimental Program

 

Materials

This study considered two sandy soils obtained from Bryan, Texas. The present research study followed US Department of Defense standard unified facilities criteria (UFC) 3-250-11 (UFC 2004, 2020) for soil stabilization with cement and polymer. Basic geo-technical tests such as wet sieve analysis, hydrometer analysis, Atterberg limits, and specific gravity tests were conducted as per respective ASTM standards (ASTM 2019). Fig. 1 shows particle size distribution curves for both soils. Soil-1 and soil-2 were classified as poorly graded sand (SP) and silty sand (SM), respectively as per the unified soil classification system (USCS). The maximum dry unit weight (MDUW) and optimum moisture content (OMC) of untreated and treated soil mixtures were determined using a modified Proctor compaction test (ASTM 2019). Table 1 presents the physical properties of both soils.

Commercially available cement type-I and vinyl acetate-ethylene copolymer were used as stabilizers for the present laboratory studies. This VAE copolymer used in the current research is a commercially available synthetic polymer and is in liquid form, has milky white color with mild odor, and is 100% dispersible. It has 55% solid content and 45% water.

 

M

σcyclic

R ¼  ϵr

ð1Þ

 

Several test methods were developed to determine the resilient modulus of unbound materials in the laboratory (Puppala 2008). However, the repeated load triaxial test is regularly used to measure the resilient modulus of unbound materials in the laboratory (Puppala 2008; Han and Vanapalli 2016). This test was designed to simulate the stress induced by traffic loading through a series of cyclic deviator stresses applied to the specimen at different confining pressures (Puppala 2008; AASHTO 2017). Past researchers observed that the resilient modulus of unbound and bound materials depends on several variables including deviator stress, confining pressure, dry density, moisture content, soil properties, stabilizer type, stabilizer dosage, curing condition, and curing time (Seed et al. 1962; Fredlund et al. 1977; Puppala et al. 1996, 2003; Drumm et al. 1997; Puppala 2008; Solanki et al. 2010; Abu-Farsakh et al. 2015; Kumar et al. 2013). Several M characterization models such as two- and three-parameter models were developed by past researchers and these models were used to characterize the resilient behavior of untreated and treated subgrade soils (Uzan 1985; Pezo 1993; Puppala et al. 2003; Puppala 2008Solanki et al. 2010; Abu-Farsakh et al. 2015; Kumar et al. 2022a). The present research was designed and focused on assessing the strength, resilient, and ductile behaviors of cement-VAE-treated sandy soils. For comparisons, both untreated and cement-treated sandy soils were also considered in this study. Extensive experimental- tal tests including modified Proctor compaction tests, unconfined compressive strength tests, and repeated load triaxial tests were conducted on both untreated and treated soils. The UCS and RLT tests were conducted at different cement and VAE dosages. Tests were

conducted on soil specimens before immersing them in the water bath and after immersing them in the water bath to investigate the moisture susceptibility of engineering properties. Two three- parameter MR characterization models were then used to analyze and model the resilient modulus test results, and regression con- stants were determined by conducting regression analysis. The ef- fect of confining pressure, deviator stress, stabilizer, and stabilizer dosages on the resilient moduli of soils were discussed in detail.

 

Experimental Program

 

Materials

This study considered two sandy soils obtained from Bryan, Texas. The present research study followed US Department of Defense standard unified facilities criteria (UFC) 3-250-11 (UFC 2004, 2020) for soil stabilization with cement and polymer. Basic geo- technical tests such as wet sieve analysis, hydrometer analysis, Atterberg limits, and specific gravity tests were conducted as per respective ASTM standards (ASTM 2019). Fig. 1 shows particle size distribution curves for both soils. Soil-1 and soil-2 were clas- sified as poorly graded sand (SP) and silty sand (SM), respectively as per the unified soil classification system (USCS). The maximum dry unit weight (MDUW) and optimum moisture content (OMC) of untreated and treated soil mixtures were determined using a modi- fied Proctor compaction test (ASTM 2019). Table 1 presents the physical properties of both soils.

Commercially available cement type-I and vinyl acetate ethyl- ene copolymer were used as stabilizers for the present laboratory studies. This VAE copolymer used in the current research is a com- mercially available synthetic polymer and is in liquid form, has milky white color with mild odor, and is 100% dispersible. It has 55% solid content and 45% water.

 

Fig. 1. Grain size distribution of soils used in this study.

 

                                                                                                             

 

 

 

Table 1. Properties of soils used in this study

 

 

Parameters                                                       Soil-1                                Soil-2

Sand (%)                                                          96.3                                   82.4

Silt (%)                                                               3.7                                   16.2

Clay (%)                                                             0.0                                      1.4

% finer than #200 sieve                                  3.7                                   17.6

USCS classification                                          SP                                      SM

Specific gravity, Gs                                           2.67                                   2.67

Plasticity index (%)                                          NP                                      NP Note: NP = nonplastic.

 

Specimen Preparation Protocols

Both control and treated soil specimens were prepared according to the US Department of Defense UFC 3-250-11 (UFC 2004, 2020) standard and AASTHO T 307-99 (AASHTO 2017) standard. Dry soil was mixed homogeneously with water for untreated/control specimens. The predetermined amount of cement was homogeneously mixed with dry soil until a uniform color was obtained and then precalculated water was thoroughly mixed with the soil-cement mixture until a uniform product was obtained for cement-treated soil specimens. For cement-VAE-treated soil specimens, the precalculated amount of VAE was then thoroughly mixed with the soil-cement-water mixture until a uniform product was achieved. The amount of water already present in the VAE was accounted for while preparing the treated soil mixtures. Each mixture was then statically compacted into a cylindrical mold having a height of 142 mm and a diameter of 71 mm at respective optimum moisture contents and maximum dry unit weights. All specimens were com- pacted within half an hour after the soil mix was prepared to prevent the initial set of soil mixtures. A sample calculation for the mass of soil, water, cement, and VAE is provided in Table 2 for cement-VAE-treated SP soil. Similar procedures and calculations were adopted for other soil specimen groups.

±

Treated soil specimens were prepared at different dosages of cement and VAE by weight of dry soil and these specimens were cured in an air-drying room (22°C ± 2°C and 50% relative humid- ity) for 7 days. Additionally, cement-treated soil specimens were prepared and cured in a moist room (22°C 2°C and 100% relative humidity) for 7 days. Table 3 summarizes stabilizer dosages and test notations followed for each soil mixture. Triplicate specimens of control and treated soils were prepared for each mixture for conducting UCS and MR tests. Overall, four sets of specimen groups were prepared, the first and second sets were prepared for dry and wet conditions for UCS tests, respectively; and the third and fourth sets were prepared for dry and wet conditions for RLT tests, respectively. Two types of control conditions have been considered for the untreated soil: one set immediately after compaction (control-1)

 

Table 2. Specimen preparation calculation for cement-VAE-treated SP soil Parameters

Table 3. Stabilizer dosages and notations used in this study

 

and the other set prepared at the same curing conditions as stabilized soil (control-2) for ensuring similar curing conditions for untreated and treated soils in this research study.
The curing condition at a temperature of 22°C 􀀁 2°C and 50% relative humidity could be considered as an air-dry process and this has been recommended over the moist curing method for polymer-treated soils (Santoni et al. 2002). Also, this air-dry curing process in the laboratory closely simulates actual field conditions during military construction procedures (Tingle and Santoni 2003; Newman and Tingle 2004; Santoni et al. 2005). Under this process, the evapora-tion of water from soil specimens hardens the polymer-soil matrix.
Unconfined Compressive Strength and Repeated Load Triaxial Tests
The UCS tests have been commonly used to evaluate the effectiveness of chemical stabilizers for road construction. The set of three specimens were cured and tested according to the dry test procedure. The remaining three specimens were immersed in the water bath for 4 h and then UCS tests were conducted as per ASTM D1633. The UCS tests were performed with a constant strain rate of 0.90% /min. The RLT tests were performed by following the AASHTO T 307-99 standard method for determining the resilient modulus of both untreated and treated soils (AASHTO 2017). All the tests were conducted using the repeated load triaxial test equipment shown in Fig. 2. These tests were conducted at three confining pressures (41.4 kPa, 27.6 kPa, and 13.8 kPa) and five deviatoric stresses (13.8 kPa, 27.6 kPa, 41.4 kPa, 55.2 kPa, and 68.9 kPa) to simulate different stress conditions induced due to traffic loading. First, the specimen was preconditioned by applying 500 load cycles with a cyclic stress of 24.8 kPa at confining pressure of 41.4 kPa. Second, the specimen was tested for 15 loading sequences that includes 100 cycles of load at different combinations of deviator stress and confining pressure. The details of the testing sequence are provided in Table 4. The preconditioning sequence helps in reducing initial irregularities between the top platen and the test specimen. Also, it helps in preventing the effects of the time interval between com-paction and loading (AASHTO 2017). The cyclic deviator stress was then applied as a haversine-shaped loading with a 0.1 s load duration, followed by a 0.9 s relaxation period, as shown in Fig. 3. The load was measured using a submersible load cell. Axial deformation of the soil specimen under repeated loading were measured using two linear variable differential transducers (LVDTs). The LVDTs were attached to the rods at 180° diametrically opposite to each other on top of the triaxial cell. Average of recoverable de-formation obtained from both LVDTs was then used in the calculation of resilient modulus.

 

Fig. 2. Repeated load triaxial test apparatus.

 

Table 4. Testing sequence for subgrade soil

 

Fig. 3. Haversine-shape and duration of each repeated load cycle.

 

using two linear variable differential transducers (LVDTs). The LVDTs were attached to the rods at 180° diametrically opposite to each other on top of the triaxial cell. Average of recoverable de-formation obtained from both LVDTs was then used in the calculation of resilient modulus.

Test Results Analysis and Discussion

Compaction Test Results
The modified Proctor compaction test results of untreated and treated SP and SM soils are shown in Figs. 4 and 5, respectively. Maximum dry unit weights and the optimum moisture contents for both untreated and treated soils are summarized in Table 5. An increase in the dry unit weight of SP soil was observed after cement treatment. Only a minor increase in the maximum dry unit weight of SM soil was observed after cement stabilization. The reason for this increase might be due to the higher specific gravity of cement relative to soils. Another reason for the increase could be the size of cement particles. Cement consists of very fine-grained particles

 

Fig. 4. Compaction curves for untreated and treated SP soil. ZAV = zero air void; and S = degree of saturation.

Fig. 5. Compaction curves for untreated and treated SM soil. ZAV = zero air void; and S = degree of saturation.

Table 5. Maximum dry unit weights and optimum moisture contents for untreated and treated soils

 

which fill the voids between sand particles which increases the dry unit weight of sand mixtures (Al-Aghbari et al. 2009). The addition of VAE to the soil-cement mixture further increased the dry unit weight of both SP and SM soils after treatment. It can be argued that the VAE copolymer provided a lubrication effect on soil particles which increased the dry unit weight of soils. Both cement and cement-VAE treatments decreased the OMC of both SP and SM soils; however, the decrease in OMC was more after the cement-VAE treatment compared to the cement treatment. The degree of saturation corresponding to OMC varied between 50% to 57% for SP soil and 62% to 65% for SM soil.
Unconfined Compressive Strength (UCS) Test Results
Fig. 6 presents the UCS test results of control and treated SP and SM soils. Tests were conducted before and after immersion of the soil specimen in the water bath. The UCS of the SP and SM control-1 soil specimens were observed to be 0.01 and 0.04 MPa, respectively. It can be observed that the UCS of untreated SP and SM soils (control-2) increased after 7 days of air-dry curing when compared to control-1 soil. This increase was because of enhancement in strength due to evaporation of moisture in the air-dry curing condition. The UCS of control-2 SM soil was higher than that of control-2 SP soil and authors attribute to higher percent finer particles in SM soil which contributed to higher matric suction during air-dry curing process. Previously, researchers studied the concept of soil suction and observed that soil matric suction increases with a reduction in water content (Fredlund and Rahardjo 1993). This concept of soil suction was used to explain results. Both control-1 and control-2 soil specimens disintegrated immediately after immersing them in water.
This study also conducted UCS tests on moist cured cement-treated specimens as per the ASTM D1632 method and these results are used here for comparison purposes. It was observed that moist-cured cement-treated specimens showed higher UCS values than air-dry cured specimens. This variation was attributed to the unavailability of water for the hydration reactions in the air-dry cured cement-treated soil specimens. UCS test results showed that cement-treated SP and SM soil specimens observed higher UCS values when compared to control soils. The UCS of cement-treated soils reduced after immersion in the water bath and the reduction in UCS values was more in the case of air-dry cured cement-treated specimens than moist cured cement-treated specimens.
UCS test results indicated a significant increase in UCS value when VAE copolymer was combined with cement treatment.

Fig. 6. Unconfined compressive strength (UCS) results for untreated and treated soils before and after immersion in water bath (a) SP soil; and (b) SM soil.

Fig. 7. Stress-strain curves for untreated and treated (a) SP; and (b) SM soils before immersion in water bath.

 

This increase in UCS value can be attributed to two factors: an increase in strength due to semi-dry conditions from the evaporation of moisture in the air-dry curing room and cementitious reactions or/and hardening of the additives within the soil matrix. The UCS values of SP+C3+P5-AD and SP+C5+P5-AD were observed to be 3.70 MPa and 4.59 MPa, respectively, and these were reduced to 0.33 and 1.02 MPa, respectively, after water bath immersion. It is important to note that the percent reduction in UCS values was more in the case of cement-VAE-treated soil specimens as compared to cement-treated soil specimens. However, the UCS values of SP+C3+P5-AD and SP+C5+P5-AD were higher than SP-C3-AD and SP-C5-AD, respectively, after water bath immersion. The probable reason for the higher loss of UCS in the case of cement-VAE treatment compared to cement treatment can be explained. Cement performed better in binding soil particles compared to VAE copolymer during moisture exposure, which resulted in this observation. In both cases, strength loss was also observed due to a change in state from semi-dry to moist after water bath immersion. SM soil also showed similar UCS trends as those of SP soil and these results can be seen in Fig. 6(b).
Figs. 7(a and b) show stress-strain curves obtained from UCS tests for untreated and treated SP and SM soil specimens, respectively. In these figures, specimens tested in dry condition (prior to immersion in the water bath) are presented. Table 6 presents a summary of axial strains at the failure of untreated and treated SP and SM soils. The axial strains at the failure have substantially increased after adding VAE to cement-treatments of SP and SM soil specimens. This indicates an increase in the ductility behavior of cement-treated soil specimens after mixing with VAE copolymer. It can be inferred that the addition of VAE to cement treatment has resulted in the semiductility behavior during axial loading when compared to cement-treated soils. This positive outcome may potentially reduce the shrinkage cracking from stiffer cement-treated soil layers on the pavement surfaces in the real field conditions as cement-treated soils often experience and induce brittle cracking in the field conditions.
Repeated Load Triaxial (RLT) Test Results
The RLT tests were conducted on both untreated and treated soil specimens before and after immersing them in the water bath. The RLT tests were conducted on triplicate soil specimens and average values were used in the plots and discussions as follows. Figs. 8 and 9 present stress and strain response of untreated SP soil under repeated load, respectively. It can be observed that the cyclic load was applied in Haversine load form for 0.1 s. The axial strain response was used to determine resilient strain, which was used in the calculation of resilient modulus. Typical variations of resilient moduli of untreated and treated SP and SM soils with various stress conditions are presented in Figs. 10 and 11, respectively. It can be observed that both control and treated SP and SM specimens showed an increase in resilient moduli with an increase in confining pressure. Control-1 SP and SM soil specimens showed an increase in resilient moduli with an increase in deviator stress. These results indicate the stress-hardening behavior of these soils when specimens are subjected to cyclic deviator stress loading. Under the stress-hardening phenomenon, the material becomes stronger with an increase in deviator stress. The resilient moduli values of control-2

 

Table 6. Strain at failure for untreated and treated soils

Fig. 8. Deviator stress response of untreated SP soil for the last 5 cycles of sequence number 5 of AASHTO T 307-99.

Fig. 9. Axial strain response of untreated SP soil for the last 5 cycles of sequence number 5 of AASHTO T 307-99.

 

SP and SM soil specimens were higher than control-1 specimens of same soils, which can be attributed to the semi-drying conditions of control-2 specimens resulting in higher moduli.
Cement-treated SP and SM soils showed stress-hardening behavior at both 3% and 5% cement dosages in both curing conditions, both before and after immersions in the water bath. An increase in resilient modulus was also observed with an increase in cement dosage. The resilient moduli of cement-treated soil specimens were reduced after immersing in the water bath at both dosages, and the reduction in the case of 3% cement treated soil was more than those at 5% cement content. This is expected because higher cement dosage results in stronger specimens with higher cementation bonding between soil particles. A significant reduction in resilient moduli of cement-treated soils cured in air-dry conditions was observed after immersion in water. Conversely, a slight reduction in the resilient moduli of cement-treated soils was observed when cured in a moist curing room. This observation is due to the lack of water for facilitating chemical reactions in air-dry curing conditions when compared to moist curing conditions. Also, loss of soil matric suction induced during air-dry curing contributed to this moduli variation.
The resilient moduli of cement-VAE-treated SP and SM soil specimens showed stress-hardening behavior at both cement dosages, before and after immersing them in the water bath. As expected, an increase in resilient modulus was observed with an increase in cement dosages for both soils. A substantial reduction in resilient moduli was observed after soaking these specimens in water and this reduction was expected and is attributed to loss of soil matric suction during saturation under immersion.
Figs. 12 and 13 present a comparison between the resilient moduli of untreated and treated SP and SM soil specimens, respectively, at confining pressure of 13.8 kPa and deviator stress of 41.4 kPa (i.e., 13th sequence in AASHTO T 307-99 test method). This stress condition was considered as it was recommended by previous studies to determine design resilient modulus value for flexible pavements (Jones and Witczak 1977; Solanki et al. 2010). It can be observed that the control-1 soil specimen showed the lowest resilient modulus for both soils. Resilient modulus of cement-VAE-treated SP soil was observed to be higher than control soils before immersion in the water bath. The SP soil treated with a mixture of 5% cement and 5% VAE showed the highest resilient modulus. Results showed that the control-2 SM soil specimen showed the highest resilient modulus and remained almost same after treatment with 5% cement and 5% VAE before immersion in the water bath. Control soil specimens disintegrated completely after immersion in the water bath, therefore, 100% reduction was observed in resilient modulus of both control soils. A substantial reduction was observed after immersing cement-VAE-treated soil specimens under water. As expected, this percent reduction was lower at higher cement

Fig. 10. Resilient modulus of untreated and treated SP soil before and after immersion in water bath (a) control-1; (b) control-2; (c) 3% cement;(d) 5% cement; (e) 3% cement þ5% polymer; and (f) 5% cement þ5% polymer.

Fig. 11. Resilient modulus of untreated and treated SM soil before and after immersion in water bath (a) control-1; (b) control-2; (c) 3% cement;(d) 5% cement; (e) 3% cement þ5% polymer; and (f) 5% cement þ5% polymer.

Fig. 12. Resilient modulus of 13th sequence for untreated and treated SP soil.

Fig. 13. Resilient modulus of 13th sequence for untreated and treated SM soil.

 

dosage treatment. The retained resilient modulus of cement-VAE-treated soils after moisture conditioning is still high indicating the need for chemical treatment to mitigate moisture susceptibility of soils. Overall, it can be mentioned that the cement-VAE treatment has not significantly resulted in the resilient property improvements of the present sandy soils.

Modeling of Resilient Moduli of Treated Soils
Different types of MR characterization models such as two-parameter and three-parameter models were developed for characterizing unbound materials (Seed et al. 1962; Uzan 1985; Pezo 1993; Mohammad et al. 1999; Ooi et al. 2004; Puppala 2008). The two-parameter models have some limitations and are highlighted by several researchers (e.g., Uzan 1985; Mohammad et al. 1999; Puppala 2008). To overcome the limitations of the two-parameter models, three-parameter models were developed. Although these models were developed for characterizing unbound materials, several researchers used them for characterizing stabilized materials (Solanki et al. 2010; Abu-Farsakh et al. 2015; Bhuvaneshwari et al. 2019; Jose et al. 2021; Kumar et al. 2022a). The current research study considered two three-parameter models developed by Pezo (1993) and National Cooperative Highway Research Program (NCHRP) project 1-28A (AASHTO 2004), mathematical equations are provided in Eqs. (2) and (4), respectively. These models were used for characterizing MR test results of both untreated and treated soils. Pezo (1993) considered both confining pressure and deviator stress in the model. The NCHRP project 1-28A developed model (also called the three-parameter universal model or M-EPDG model) accounts for bulk stress and octahedral shear stress. The M-EPDG recommends the three-parameter universal model for characterizing unbound materials

fx

where, MR = resilient modulus; k1p, k2p and k3p = regression con-stants for Pezo (1993) model; k1m, k2m and k3m = regression constants for M-EPDG model; σd = deviator stress = σ1 − σ3; θ = bulk stress = σ1 þ σ2 þ σ3; Pa = atmospheric pressure; and τoct = octahedral shear stress.

fx

 

All three model constants were determined by conducting multiple-linear regression analyses on the resilient moduli test results. Tables 7 and 8 present regression constants and the coefficient of determination (R2) for both untreated and treated SP and SM soils, respectively. This research study obtained a minimum value of R2 of 0.94 for SP soil and 0.92 for SM soil, which indicates an excellent fit between the model predictions and experimental test results.
Based on the results presented in Tables 7 and 8, the following observations are noted:

  • The model constant, k1p, increased after cement-VAE treatment (in dry test condition) as compared to control-2 specimen for SP soil. On the contrary, this constant was decreased after cement-VAE treatment compared to control-2 specimen for SM soil. It was observed that the value of k1p reduced after immersing in the water bath for both soils. Similar trends as k1p were observed for k1m for both soils.
  • The coefficient, k2p, which describes the stiffening behavior of the treated material with an increase in confining pressure, was less than 1.0 for all untreated and treated SP and SM soils, and these values varied between 0.125 and 0.531. This indicates that the effect of confining pressure decreases with an increase in resilient modulus.
  • The value of k2p decreased after air-drying control soil specimens (control-2) as compared to control-1 specimens for both soils. The cement-VAE treatment (in dry test condition) further reduced the value of k2p as compared to control-2 specimens. This observation shows that the effect of confining pressure on MR decreased after air-drying or treatment. It was observed that the value of k2p increased after immersing cement-VAE-treated soil specimens in the water bath compared to that of before immersion for both soils. Both characterization models showed similar behavior of the k2 parameter for all soil groups.

Table 7. Three-parameter model constants for control and treated SP soils

Table 8. Three-parameter model constants for control and treated SM soils

 

Fig. 14. Measured MR versus predicted MR we (a) Pezo model (Pezo 1993); and (b) M-EPDG model.

 

  • In the Pezo (1993) model, a positive and negative value of the parameter, k3p, represents stress-hardening and stress-softening behavior, respectively. The value of k3p was positive for both soils and it varied between 0.016 and 0.234, which shows stress-hardening behavior. Control-1 specimens of both SP and SM soils showed negative values of k3m for the M-EPDG model. It was observed that the M-EPDG model showed both positive and negative values of k3m for treated SP and SM soil specimens.
  • The coefficient k3p decreased after cement-VAE treatment (in dry test condition) compared to control-2 specimen for both soils and further reduced after immersing in the water bath. It was observed that the value of k3m decreased after 3% cement and 5%VAE treatment and increased after 5% cement and 5% VAE treatment as compared to control-2 specimens for both soils, and these values reduced after immersing in the water bath.
    The predicted resilient modulus values were compared with the measured resilient modulus values, as shown in Figs. 14 and 15 for both SP and SM soils, respectively. Both Pezo (1993) and M-EPDG models reasonably characterized and predicted the resilient modulus values in the back-calculation analysis.

Conclusions and Recommendations
A research study presented in this paper has comprehensively assessed the effectiveness of the mixture of cement and VAE in enhancing the unconfined compressive strength and resilient moduli of sandy soils while reducing the brittle behavior that is usually associated with cement-treated soils. Various variables including cement dosages and curing conditions were studied and addressed. Both UCS and RLT tests were conducted on untreated and treated soils compacted at optimum conditions from modified Proctor test results. The resilient modulus test results were analyzed using three-parameter models such as Pezo (1993) and M-EPDG models. The following conclusions are drawn based on the experimental study results:

  1. Cement and cement-VAE treatments of SP and SM soils have increased the maximum dry unit weight or MDUW compared to untreated soils. The increase in MDUW was more in the case of SP soil as compared to SM soil after cement and cement-VAE stabilizations. The optimum moisture content was reduced slightly after cement and cement-VAE treatments for both soils.
  2. Both cement and cement-VAE treatments improved the unconfined compressive strength properties of both SP and SM soils. The UCS results of cement-treated specimens cured in a moist room were higher than those soils cured in the air-drying room. This was due to the lack of sufficient water for cement hydration reaction when specimens were cured in the air-drying room. An increase in UCS was, however, observed with an increase in cement dosages.
  3. A significant increase in the UCS values of SP and SM soils was reported after cement-VAE treatment. Also, the UCS value was increased with an increase in cement percentage in the cement-VAE mixture. This increase in strength was provided due to a combination of the development of soil matric suction induced due to the evaporation of moisture in soil specimens and the bonding between soil particles created by the mixture of cement and VAE. The UCS values were reduced after immersing treated soil specimens under water. The addition of VAE into cement increased the axial strain at the failure which shows an improvement in ductile behavior after the combined cement and polymer stabilization method as compared to cement treatment alone.
  4. The resilient moduli values of control-2 SP and SM soil specimens were higher than control-1 specimens of same soil, which can be attributed to the semi-drying conditions of control-2 specimens resulting in higher moduli. Control soil specimens disintegrated immediately after immersion in the water bath. Resilient modulus of cement-VAE-treated SP soil was observed to be higher than control soils before immersion in the water bath. The control-2 SM soil specimen showed the highest resilient modulus and remained almost the same after treatment with 5% cement and 5% VAE. A substantial reduction was observed after immersing these cement-VAE-treated soil specimens under water. The retained resilient modulus of cement-VAE-treated soils after moisture conditioning is still high when compared to control-2 soils indicating the role of chemical treatment to mitigate moisture susceptibility of these soils.
  5. Both three-parameter models provided a higher coefficient of determination and obtained an excellent fit with experimental data. Therefore, it is recommended to use both three-parameter models for the determination of the resilient modulus of soils considered in this research study.

The model parameters, k1p and k1m, of cement-VAE-treated SP and SM soils were higher and lower than the same control-2 soil specimens, respectively. The coefficients k2p and k2m decreased after cement-VAE treatment compared to the control specimens for both soils. This shows that the effect of confining pressure or bulk stress on the resilient modulus is reduced after treatment. Both cement-VAE-treated soils showed lower values of k3p coefficient compared to control-2 specimen. No definite trend was observed in the case of k3m coefficient. These k3p and k3m values reduced after immersing in the water bath for both soils.
The present research study observed an improvement of UCS and resilient moduli of SP and SM soils after the treatment with a mixture of cement and VAE and moduli improvements are moderate with the addition of VAE. Furthermore, the VAE additive enhanced the failure behavior of cement-treated specimens from brittle to semi-ductile. Therefore, the mixture of cement and polymer can be recommended for the stabilization of soils considered in this study. Both Pezo (1993) and M-EPDG models precisely predicted the resilient modulus values and the authors recommend these models for soils with characteristics similar to the soils used in the current study.

Data Availability Statement

All data generated or used during the study appear in the form of figures and tables in the published article. Some or all data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments
The authors would like to acknowledge US Army Engineer Research and Development Center, Vicksburg, Mississippi for granting the funds for Research project #W912HZ 20P0090. Also, the authors would also like to acknowledge Mr. Zoheb Faisal, a graduate student for his help during experimental works.

References
AASHTO. 2004. Mechanistic-empirical pavement design guide. Washington, DC: AASHTO.

AASHTO. 2017. Standard method of test for determining the resilient modulus of soils and aggregate materials. AASHTO T 307-99. Washington, DC: AASHTO.

Abu-Farsakh, M., S. Dhakal, and Q. Chen. 2015. “Laboratory characterization of cementitiously treated/stabilized very weak subgrade soil under cyclic loading.” Soils Found. 55 (3): 504–516. https://doi.org/10.1016/j .sandf.2015.04.003.

Al-Aghbari, M. Y., Y. E. A. Mohamedzein, and R. Taha. 2009. “Stabilisation of desert sands using cement and cement dust.” Proc. Inst. Civ. Eng. Ground Improv. 162 (3): 145–151. https://doi.org/10.1680/grim .2009.162.3.145.

ASTM. 2019. Annual book of ASTM standards. West Conshohocken, PA:ASTM.

Behnood, A. 2018. “Soil and clay stabilization with calcium- and non-calcium-based additives: A state-of-the-art review of challenges, approaches and techniques.” Transp. Geotech. 17 (Dec): 14–32. https://doi.org/10.1016/j.trgeo.2018.08.002.

Bhuvaneshwari, S., R. G. Robinson, and S. R. Gandhi. 2019. “Resilient modulus of lime treated expansive soil.” Geotech. Geol. Eng. 37 (1): 305–315. https://doi.org/10.1007/s10706-018-0610-z.

Drumm, E. C., J. S. Reeves, M. R. Madgett, and W. D. Trolinger. 1997. “Subgrade resilient modulus correction for saturation effects.” J. Geotech. Geoenviron. Eng. 123 (7): 663–670. https://doi.org/10.1061/(ASCE) 1090-0241(1997)123:7(663).

Fredlund, D. G., A. T. Bergan, and P. K. Wong. 1977. “Relation between resilient modulus and stress research conditions for cohesive subgrade soils.” Transp. Res. Rec. 642 (Mar): 73–81.

Fredlund, D. G., and H. Rahardjo. 1993. Soil mechanics for unsaturated soils. New York: Wiley.

Georgees, R. N., R. A. Hassan, R. P. Evans, and P. Jegatheesan. 2017. “Resilient response characterization of pavement foundation materials using a polyacrylamide-based stabilizer.” J. Mater. Civ. Eng. 30 (1): 04017252. https://doi.org/10.1061/(ASCE)MT.1943-5533.0002109.

Han, Z., and S. K. Vanapalli. 2016. “State-of-the-art: Prediction of resilient modulus of unsaturated subgrade soils.” Int. J. Geomech. 16 (4):04015104. https://doi.org/10.1061/(ASCE)GM.1943-5622.0000631.

Huang, J., R. B. Kogbara, N. Hariharan, E. A. Masad, and D. N. Little. 2021. “A state-of-the-art review of polymers used in soil stabi-lization.” Constr. Build. Mater. 305 (Oct): 124685. https://doi.org/10 .1016/j.conbuildmat.2021.124685.

Iyengar, S. R., E. Masad, A. K. Rodriguez, H. S. Bazzi, D. Little, and H. J. M. Hanley. 2012. “Pavement subgrade stabilization using poly-mers: Characterization and performance.” J. Mater. Civ. Eng. 25 (4): 472–483. https://doi.org/10.1061/(ASCE)MT.1943-5533.0000612.

Jones, M. P., and M. W. Witczak. 1977. “Subgrade modulus of the San Diego test road.” Transp. Res. Rec. 641: 1–6.

Jose, A., J. M. Krishnan, and R. G. Robinson. 2021. “Resilient and permanent deformation response of cement-stabilized pond ash.” J. Mater. Civ. Eng. 34 (1): 1–14. https://doi.org/10.1061/(ASCE)MT.1943-5533 .0004044.

Kolay, P. K., B. Dhakal, S. Kumar, and V. K. Puri. 2016. “Effect of liquid acrylic polymer on geotechnical properties of fine-grained soils.” Int. J. Geosynth. Ground Eng. 2 (4): 1–9. https://doi.org/10.1007/s40891-016-0071-5.

Kumar, P., A. J. Puppala, S. S. C. Congress, K. Ramineni, and J. S. Tingle. 2013. “Influence of compaction characteristics and moisture exposure on resilient moduli of cement-treated soil.” In Geo-Congress 2023: Soil Improvement, Geoenvironmental, and Sustainability, Geotechnical Special Publication 339, edited by E. Rathje, B. M. Montoya, and M. H. Wayne, 370–380. Reston, VA: ASCE.

Kumar, P., A. J. Puppala, J. S. Tingle, S. Chakraborty, S. Sarat, and C. Congress. 2022a. “Resilient characteristics of polymer emulsion-treated sandy soil.” Transp. Res. Rec. 2676 (9): 1–13. https://doi.org/10.1177 /03611981221087238.

Kumar, P., J. S. Tingle, A. J. Puppala, S. Chakraborty, and S. S. C. Congress. 2022b. “Permeability characteristics of polymer emulsion-stabilized soils.” In Vol. 331 of Geo-Congress 2022, Geotechnical Special Publication 331, edited by A. Lemnitzer and A. W. Stuedlein, 261–270. Reston, VA: ASCE.

Little, D., and S. Nair. 2009. “Recommended practice for stabilization of subgrade soils and base materials.” In National cooperative highway research program. Washington, DC: Transportation Research Board.

Little, D. N., S. Nair, and B. Herbert. 2009. “Addressing sulfate-induced heave in lime treated soils.” J. Geotech. Geoenviron. Eng. 136 (1): 110–118. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000185.

Mohammad, L. N., B. Huang, A. J. Puppala, and A. Allen. 1999. “Regression model for resilient modulus of subgrade soils.” Transp. Res. Rec. 1687 (1): 47–54. https://doi.org/10.3141/1687-06.

Newman, K., and J. S. Tingle. 2004. “Emulsion polymers for soil stabilization.” In Proc., 2004 FAA Worldwide Airport Technology Transfer Conf. Washington, DC: Federal Aviation Administration.

Onyejekwe, S., and G. S. Ghataora. 2015. “Stabilization of quarry fines using a polymeric additive and portland cement.” J. Mater. Civ. Eng. 28 (1): 04015070. https://doi.org/10.1061/(ASCE)MT.1943-5533 .0001324.

Ooi, P. S. K., A. R. Archilla, and K. G. Sandefur. 2004. “Resilient modulus models for compacted cohesive soils.” Transp. Res. Rec. 1874 (1): 115–124. https://doi.org/10.3141/1874-13.

Papagiannakis, A. T., and E. A. Masad. 2008. Pavement design and materials. New York: Wiley.

Petry, T. M., and D. N. Little. 2002. “Review of stabilization of clays and expansive soils in pavements and lightly loaded structures—History, practice, and future.” J. Mater. Civ. Eng. 14 (6): 447–460. https://doi.org /10.1061/(ASCE)0899-1561(2002)14:6(447).

Pezo, R. 1993. “A general method of reporting resilient modulus tests of soils: A pavement engineer’s point of view.” In Proc., 72nd Annual Meeting of the Transportation Research Board. Washington, DC: Transportation Research Board.

Prusinski, J. R., and S. Bhattacharja. 1999. “Effectiveness of Portland cement and lime in stabilizing clay soils.” Transp. Res. Rec. 1652 (1): 215–227. https://doi.org/10.3141/1652-28.

Puppala, A., L. Mohammad, and A. Allen. 1996. “Engineering behavior of lime-treated Louisiana subgrade soil.” Transp. Res. Rec. 1546 (1): 24–31. https://doi.org/10.3141/1546-03.

Puppala, A. J. 2008. Estimating stiffness of subgrade and unbound materials for pavement design. Washington, DC: Transportation Research Board.

Puppala, A. J. 2016. “Advances in ground modification with chemical additives: From theory to practice.” Transp. Geotech. 9 (Dec): 123–138. https://doi.org/10.1016/j.trgeo.2016.08.004.

Puppala, A. J. 2021. “Performance evaluation of infrastructure on problematic expansive soils: Characterization challenges, innovative stabilization designs, and monitoring methods.” J. Geotech. Geoenviron. Eng. 147 (8): 04021053. https://doi.org/10.1061/(ASCE)GT.1943-5606.0002518.

Puppala, A. J., N. Intharasombat, and R. K. Vempati. 2005. “Experimental studies on ettringite-induced heaving in soils.” J. Geotech. Geoenviron. Eng. 131 (3): 325–337. https://doi.org/10.1061/(ASCE)1090-0241 (2005)131:3(325).

Puppala, A. J., A. M. Ramakrishna, and L. R. Hoyos. 2003. “Resilient moduli of treated clays from repeated load triaxial test.” Transp. Res. Rec. 1821 (May): 68–74. https://doi.org/10.3141/1821-08.

Rezaeimalek, S., R. Nasouri, J. Huang, and S. Bin-Shafique. 2018. “Curing method and mix design evaluation of a styrene-acrylic based liquid polymer for sand and clay stabilization.” J. Mater. Civ. Eng. 30 (9): 04018200. https://doi.org/10.1061/(ASCE)MT.1943-5533.0002396.

Santoni, R. L., J. S. Tingle, and M. Nieves. 2005. “Accelerated strength improvement of silty sand with nontraditional additives.” Transp. Res. Rec. 1936 (1): 34–42. https://doi.org/10.1177/0361198105193600105.

Santoni, R. L., J. S. Tingle, and S. L. Webster. 2002. “Stabilization of silty sand with nontraditional additives.” Transp. Res. Rec. 1787 (1): 61–70. https://doi.org/10.3141/1787-07.

Seed, H., C. Chan, and C. E. Lee. 1962. “Resilience characteristics of subgrade soils and their relation to fatigue failures in asphalt pavements.” In Proc., Int. Conf. on the Structural Design of Asphalt Pavements, 611–636. Washington, DC: Transportation Research Board.

Solanki, P., M. M. Zaman, and J. Dean. 2010. “Resilient modulus of clay subgrades stabilized with lime, class c fly ash, and cement kiln dust for pavement design.” Transp. Res. Rec. 2186 (1): 101–110. https://doi.org /10.3141/2186-11.

Tingle, J. S., K. Newman, S. L. Larson, C. A. Weiss, and J. F. Rushing. 2007. “Stabilization mechanisms of non-traditional additives.” Transp. Res. Rec. 1989 (1): 59–67. https://doi.org/10.3141/1989-49.

Tingle, J. S., and R. L. Santoni. 2003. “Stabilization of clay soils with non-traditional additives.” Transp. Res. Rec. 1819 (1): 72–84. https://doi.org /10.3141/1819b-10.

UFC (Unified Facilities Criteria). 2004. Soil stabilization for pavements. UFC 3-250-11. Washington, DC: Dept. of Defense.

UFC (Unified Facilities Criteria). 2020. Soil stabilization and modification for pavements. UFC 3-250-11. Washington, DC: Dept. of Defense.

Uzan, J. 1985. “Characterization of granular material.” Transp. Res. Rec. 1022 (1): 52–59.

Yaowarat, T., S. Horpibulsuk, A. Arulrajah, F. Maghool, M. Mirzababaei, A. S. A. Rashid, and A. Chinkulkijniwat. 2022. “Cement stabilisation of recycled concrete aggregate modified with polyvinyl alcohol.” Int. J. Pavement Eng. 23 (2): 349–357. https://doi.org/10.1080/10298436 .2020.1746311.

Zhang,B.,W.Jiang,Q.Xu,D.Yuan,J.Shan,andR.Lu.2022. “Experimental feasibility study of ethylene-vinyl acetate copolymer (EVA) as cement stabilized soil curing agent.” Road Mater. Pavement Des. 23 (3): 617–638. https://doi.org/10.1080/14680629.2020 .1834442.

Complete the form below to download this document now.

https://bit.ly/3OVQj5d