An Experimental Study on Unconfined Compressive Strength of Soft Soil-Cement Mixtures with or without GGBFS in the Coastal Area of Vietnam

Soft soil is widely distributed in Vietnam, especially in the coastal area. In engineering practice, soft soil cannot be used to build any construction and needs to be improved or treated before building construction. In addition, Vietnam has many pig-iron or thermal power plants, which annually produce a huge amount of granulated blast furnace slag (GBFS). (us, the use of this material for soft soil improvement needs to be considered. (is paper presents experimental results on the unconfined compressive strength (UCS) of three Vietnam’s soft soils treated with Portland cement and Portland cement with ground granulated blast furnace slag (GGBFS). Binder dosage used in this study is 250, 300, and 350 kg/m with the three different water/cement ratios of 0.8, 0.9, and 1.0, respectively. (e research results showed that the UCS of soil-cement mixtures depends on soil type, water/cement ratio, cement type, and binder content. Accordingly, the unconfined compressive strength increased with the increase of binder contents, the decrease of the natural water content of soft soil, water/cement ratios, and clay content. (e highest value of UCS of treated soils was found for the soil at Site II with the Portland cement content, cement GGBFS, and water/ cement ratio of 873 kg/m, 2355 kg/m, and 0.8, respectively. Besides, for all the three soils and two binder types, the water/cement ratio of 0.8 was found to be suitable to reach the highest UCS values of treated soil. (e research results also showed that the UCS of treated soil with cement GGBFS was higher than that of treated soil with Portland cement.(is indicated the effectiveness of the use of Portland cement with GGBFS in soft soil improvement. (ere is great potential for reducing the environmental problems regarding the waste materials from pig-iron plants in Vietnam and the construction cost as well.


Introduction
Vietnam is one of the countries with a long coastline of more than 3000 km [1]. Plains are distributed along the coastline and formed by sedimentation from many rivers such as Red, ai Binh, Ma, Dong Nai, and Mekong. e stratigraphy of these deltas is very complicated and is formed in the Holocene age, that is, young sediments. ese sediments are mainly in soft to very soft state and often make an unfavorable condition for construction activities [1][2][3][4][5][6]. Recently, the demand for infrastructure development in Vietnam is on the rise. us, there is great attention to the areas with poor soil conditions. However, the soil in these areas needs to be improved before building construction. So far, one of the methods that are commonly used for the improvement of soft soil in Vietnam is cement deep mixing (CDM) method [7][8][9][10][11]. is method has been also widely used in the world, especially in southeast Asian countries [12].
Regarding the use of cement for soil improvement, some main properties of soil-cement mixtures, including the unconfined compressive strength, the elastic modulus (E 50 ), small-strain shear modulus, and strain-dependent shear stiffness, have been widely investigated [12][13][14][15][16][17][18][19]. e unconfined compressive strength (UCS) is one of the most important strength parameters of cement-stabilized soils and has received much attention from the literature [15][16][17][20][21][22][23][24][25]. Abbey et al. [24] indicated that UCS is the most essential parameter to design the cut-off walls for the prevention of seepage in water-retaining walls. It was also found that the UCS was an indication of the degree of reaction in the soil-binder-water mixture based on the rate of hardening of the improved mixture. e UCS of soil-cement mixtures is often affected by many factors such as soil properties, the chemical reaction, cementitious constituent [26], and type of mixing method [27]. e development of strength of soil treatment depends on the mineralogy of soils, types and cement content, curing period, and environmental conditions of soft ground [12]. Kitazume and Tersahi [28] indicated that the most important factors that affect the properties of soil-cement columns are the binder type, soil properties, mixing type, and curing conditions.
In general, the CDM method is an economical and effective method for soft soil improvement. is method can significantly reduce the compressibility and permeability of soil and increases the soil bearing capacity and the resistance against liquefaction [25]. However, one of the main limitations of using cement for soil improvement is that the production of cement often consumes a big amount of heating energy and can release a huge amount of carbon dioxide (CO 2 ), which results in global warming [29,30]. Now, there is an increasing trend of using waste materials such as granulated blast furnace slag (GBFS), ground granulated blast furnace slag (GGBFS), fly ash, bottom ash, and steel slag, which can cause a serious problem for the environment in civil engineering practice. e utilization of these materials can reduce the negative impact on the environment and replace the natural materials such as sand, gravel, or even conventional Portland cement. GBFS is an industrial by-product, and production of GBFS consumes less energy and releases less CO2 emissions rather than the production of Portland cement [29,31,32]. GGBFS, a ground product of grinding GBFS, often contains a high content of calcium oxide and silica, which can react with alkali as a pozzolanic reaction to form the cementitious products [33,34]. us, GGBFS has been widely investigated to partly replace cement in concrete products [35][36][37][38] and soil stabilization. For soil stabilization, the GGBFS is often combined with cement, lime for stabilization of soft soil, lateritic soils, and expansive soils for the base and subbase layers in road and pavement applications [34,[39][40][41][42]. However, the research on the utilization of GGBFS to partly replace cement in soil improvement using the CDM method is still limited.
In Vietnam, it is predicted that the annual production of GBFS from iron-pig plants in the whole country will be above 7 million tons in 2020 [43]. is waste material often makes a serious problem for soil, water, and air environments in many areas where iron plants are distributed. Our previous study showed that GBFS has the same effect as the coal ash from thermal power plants on the environment in Vietnam [44]. us, the amount of GBFS must be reduced for environment management. In Vietnam, GBFS can be reused as base material for rural roads [45]. However, there is still a huge amount of GBFS and there is a great potential of using GGBFS, which is a ground product of GBFS, in combination with cement for soil improvement. As mentioned above, the cement-stabilized soil, especially the CDM method, is very necessary and popular in Vietnam. Hence, the combination of GGBFS with cement for soft soil improvement using deep cement mixing method is very practical in Vietnam. In this study, Portland cement and GGBFS will be used to mix with some types of soft soil distributed along the Vietnamese coastal area to investigate the applicability of GGBFS in soft soil treatment. In addition, some factors affecting the unconfined compressive strength of soil-cement and soil-cement GGBFS mixtures such as types of soil, binder content, and water/cement ratios will be evaluated. Accordingly, a number of remolded samples of soil-cement and soil-cement GGBFS were prepared and tested in the laboratory.

Materials.
e soil samples were taken in boreholes from different locations in the coastal area of Vietnam, including anh Hoa province (Site I), Nghe An province (Site II), and Ho Chi Minh city (Site III). e soil samples were first determined the physical properties, unconfined compressive strength, and particle size analysis. e tests were conducted in accordance with the ASTM standards. According to the Unified Soil Classification System (USCS), three soils are classified as CH-Fat clay, SC-Clayey sand, and CH-Fat clay, respectively. Some properties of these soils are provided in Table 1, and the particle size distribution is shown in Figure 1. It can be seen that these soils have high water content and void ratio. e natural water contents of soft clay in Site I and Site III are higher than those in Site II. ese soils are in a soft state and their strengths need to be improved to meet the construction requirements.
e cement used in this study was Portland cement (PCB 40) and Portland cement with ground granulated furnace blast (PCB BFS 40) made by Ha Tien company. e chemical compositions of these types of cement are shown in Table 2. PCB BFS 40 has 40∼50% GGBFS content.
Some properties of PCB40 and PCB BFS 40 are shown in Table 3.

Investigating the Unconfined Compressive Strength of
Soil-Cement-GGBFS Mixture. In this study, the effects of soil types, cement types, cement contents, and water/cement ratios on the UCS values of treated soil will be investigated. To investigate the effect of soil type, three different types of soft soil as mentioned above were used to mix with cement PCB40 and cement PCB BFS 40. From there, the effects of two types of cement (Portland cement and Portland cement with GGBFS) on the UCS were also evaluated.
To study the effect of cement content on the unconfined compressive strength of soil-cement mixtures, a group of tests were conducted on three dosage rates, which were 250, 300, and 350 kg/m 3 . e cement (cement GGBFS) proportion is defined as the weight of binder added per unit volume of the soil to be treated, expressed in kg/m 3 [46].
Variation of unconfined compressive strength under the effect of different water/cement ratios was described at 28-day curing. In this study, three different water/cement (W/ C) ratios of 0.8, 0.9, and 1.0 were investigated at three different soil types, three different binder contents, and two types of cement.

Sample Preparation and Testing.
e undisturbed soil samples were collected from the borehole at three sites. e soil samples were stored at the room humidity and temperature to retain the moisture content. e soil was then mixed with Portland cement and Portland cement with GGBFS (cement GGBFS). e mixtures of soil-binder were prepared with W/C ratios of 0.8, 0.9, and 1.0. e contents of PCB40 and PCB BFS 40 are shown in Table 4.
A metal mold with 50 mm inner diameter and 100 mm length was used for the preparation of treated samples. e treated samples were created by mixing soil, cement (cement GGBFS), and water. e amounts of Portland cement and cement GGBFS and water/cement ratios were used as shown in Table 4. Natural soil samples were firstly mixed for 5-10    minutes using a mixer; then the slurry of binder-water was poured into it and mixed again for 5-10 minutes. e slurry of mixtures was then placed into a metal mold in three layers. e surface of the sample in the mold was flattened and weighted to determine the mass. e samples were then sealed tightly using a plastic sheet to prevent loss of moisture content. All the treated samples were subjected to cure at room humidity and room temperature of 25 ± 2°C for 28 days before testing. e unconfined compressive strength tests were carried out on 108 cylindrical specimens of soft soil-Portland cement mixtures and 108 cylindrical specimens of soil-cement GGBFS mixtures. e unconfined compressive strength tests were performed on these specimens in accordance with ASTM D 2166. e rate of loading for the unconfined compressive tests was 1 mm/min.

Results and Discussion
Specimens were tested at three different binder contents (250 kg/m 3 , 300 kg/m 3 , and 350 kg/m 3 ) and three water/ cement ratios (0.8, 0.9, and 1.0). e results of unconfined compressive strength (UCS) are shown in Table 5 and In general, for all the water/cement ratios and binder contents, the highest value of UCS of treated soils is found for the soil at Site II, followed by the soil at Site I and that at Site III. Concerning the physical properties of samples used in Table 1, it can be seen that the UCS values of treated soil tend to decrease with the increase of clay content and liquid limit of the original soils. As reported by Farouk and Shahien [46], the compressive strength of the cemented soil decreased with the increase in the fines content of untreated soil. e treated soil at Site III has the lowest value of UCS, since the untreated soil has the highest clay content and the highest liquid limit. Horpibulsuk et al. [22] indicated that the clay-water/cement ratio, that is, the ratio of the initial water content of the clay (w c ) to the cement content, plays an important role in the mix design in the laboratory and in analyzing the strength and deformation behavior of soilcement mixtures. Accordingly, the increase of the claywater/cement ratio leads to a decrease in the UCS value of soil-cement mixture. Binder content plays an important role in treated soil, since it strengthens the fabric at the intercluster spacing of treated soil. In soft clay, water is already present and is not free, but it can interact with the clay [46].

Effect of Binder Content on the Unconfined Compressive
Strength.
e effects of binder content on unconfined compressive strength (UCS) are presented in Figures 5-7. e unconfined compressive strengths of cement-treated soil (S-C) and cement-GGBFS-treated soil (C-F-S) range from 648 to 951 kPa and from 1539 to 2816 kPa, respectively. e lowest value of unconfined compressive strength is found in a specimen that has the smallest cement content (250 kg/m 3 ). e highest one is observed in the specimen that has the highest cement content (350 kg/m 3 ). It can be seen that the increase of unconfined compressive strength is attributed to the increase in cement content or cement GGBFS content. is is consistent with the research result of Asturias and Lorenzo [47]. It is possible that the increase in UCS of the treated soil is due to the reaction between cement and pore water and results in cation exchange and formation of cement products that bind the soil particles together [41].
Farouk and Shahien [46] also showed that the strength, stiffness, and brittleness of the cemented soil increased with the increase of cement dose. Site I -250 Site II -250 Site III -250 Site I -300 Site II -300 Site III -300 Site I -350 Site II -350 Site III - Site II -250 Site III -250 Site I -300 Site II -300 Site III -300 Site I -350 Site II -350 Site III -350 PCB40 PCB40BFS Site II -250 Site III -250 Site I -300 Site II -300 Site III -300 Site I -350 Site II -350 Site III -350 PCB40 PCB40BPS respectively. It was believed to be due to the effect of GGBFS, which makes higher reaction than the cement without GGBFS. e presence of GGBFS containing high CaO content will increase cementitious products through a pozzolanic reaction. In addition, the size of particles may increase due to the flocculation and then results in an increase in strength [19,20]. Moreover, the increase in strength may also be due to the possibility of suction development in pore fluid because of partial saturation of the improved samples after curing [48]. e results of the present study indicated that the cement GGBFS can reduce the amount of cement used for the improvement of soft soil. Pathak et al. [49] also reported that the increase in the percentage of GGBFS could lead to considerably increasing the strength and reducing pavement thickness for improvement of the subgrade layer. Figures 11-13 illustrate the variation in unconfined compressive strength of soil-cement mixtures and soil-cement-GGBFS mixtures with different water/cement (W/C) ratios. e highest UCS corresponds to specimens having the lowest W/C ratio. e smallest UCS corresponds to the specimen that has the highest W/C ratio. It can be observed that the unconfined compressive strength, UCS, decreased with the increasing W/C ratio. With the increase of W/C from 0.8 to 1.0, the unconfined compressive strength of S-C mixtures decreased from 3.02% to 13.61%, from 6.93% to 17.21%, and from 3.16% to 10.21% at Site I, Site II, and Site III, respectively. For S-F-C mixtures, the UCS at Site I, Site II, and Site III decreased from 12.1% to 20.29%, from 3.67% to 11.1%, and from 11.89% to 21.51%, respectively, when the W/C ratios increased from 0.8 to 1.0.

Advances in Civil Engineering
It is also observed by Farouk and Shahien [46] that the compressive strength decreased with the increase in water/ cement ratio. It is also consistent with the investigation of Pathivada [50] at three different cement ratios of 0.8, 1.0, and 1.3. It is in good agreement with that achieved by Liu et al. [51] that if water content was less than 0.9, CDM (cement deep mixing) behaves better in terms of strength, if water content was within 0.9∼1.0, DJM (dry jet mixing) and CDM had a comparative effect, and if water content ratio was larger than 1.0, DJM behaved better. It was believed that the decrease of UCS is due to excess water required for cement hydration [52]. When the water/cement ratio increases, the relative cement content decreases and this increases void ratio and water content of the soil-cement [51].

Conclusions
Based on this investigation, the following observation and conclusions are made as follows. e effect of soil types on UCS was evaluated using three soils along the coastline of Vietnam and two types of cement. e research result shows that, for all the binder contents and the water/cement ratios, the highest UCS value is observed for the soil at Site II. In addition, for the three types of soil and three contents of binder, the water/cement ratio of 0.8 results in the highest UCS value of treated soil. e unconfined compressive strength of stabilized soil decreases with the increase in clay fraction and liquid limit of the original soil. Besides, it also decreases with the increase of water/cement ratios and the natural water content of clay and the decrease of binder contents. e use of Portland cement with GGBFS (cement GGBFS) is more effective than Portland cement in the improvement of soft soils.
us, the cement GGBFS has shown great potential for soft soil improvement in Vietnam.
is will help reduce both the environmental pollution and the construction cost.

Data Availability
e data used to support the findings of this study are included in the published article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.