Strength and Microfabric of Expansive Soil Improved with Rice Husk Ash and Lime

Expansive soil has harmful eﬀect on engineering. Rice husk ash (RHA) has high pozzolanic activity, so it can form new cementing material with lime or cement to solidify soil. In this paper, the tests of free expansion rate, water ratio limit, and optimum moisture content (OMC) are carried out; then, RHA and lime were added to artiﬁcial soil in diﬀerent proportions of 5, 10, 15, and 20% by weight, in which the ratio of RHA to lime is 80: 20. The unconﬁned compressive strength (UCS) in diﬀerent curing age is measured, and the improvement eﬀect of RHA and lime to expansive soil can be obtained. Finally, the reason of improvement eﬀect is explained by using the scanning electron microscope (SEM). The results of the study show that (1) for the best utilization eﬀect, the optimum percentage of RHA is 12% and lime is 3%; (2) the UCS is 2.6 times of the pure soil after curing of 14d under the optimum percentage; (3) the curing age has a signiﬁcant eﬀect on strength; (4) the main reason for the strength increase of the modiﬁed soil is that the crystal produced by the pozzolanic activity ﬁlls the pores of the soil.


Introduction
Expansive soil is a special kind of clay, which is rich in hydrophilic minerals. At present, there are different methods to solve expansive soil. In recent years, the geopolymerization method is carried out to stabilize expansive soil.
e geopolymerization occurs when an amorphous material rich in silica is mixed with soil in alkaline environment, such as bagasse ash and lime [1], fly ash and lime [2,3], and volcanic ash and lime [4]. Among them, lime, as an additive, is the most commonly used to improve expansion soil. However, lime will change the Ph of soil which can affect the environment. erefore, reducing the amount of lime is a very important subject in foundation engineering.
Rice husks account for about 20 percent of the total weight of rice. e quantity of rice husk is large, but there is no reasonable way to utilize it on a large scale. e biomass power generation technology has alleviated the problem to some extent, but the enterprises have no idea to deal with the RHA produced after the power generation. Cook [5] believes that RHA, which is rich in silicon, can be used as pozzolanic material to produce geopolymer to improve the performance of expansive soil. erefore, some scholars have added RHA into the expansive soil to study its engineering properties.
Aziz [6] found that the RHA has a great influence on the high plastic soil. It can decrease the swelling property and increase the strength. Singh [7] indicated that when the content of RHA in the soil samples is 10%, the CBR and unsoaked CBR reach their peak and the expansive force is the lowest. e engineering properties of soil mixed with RHA and its application prospect on roadbed were studied by Muntohar [8], Rao and Ganja [9], and Oviya [10]. In their studies, RHA played an important role in soil improvement. Kumar [11] found that when the lime content is more than 3%, there is no substantial effect on the strength increase. Agus Muntohar's research studies [12] showed that more than 10% lime content would cause plastic failure of soil during compression. Sabat [13] found that the curing period should be more than 7 days (d). Liu et al. [14] presented a cementitious material combined with RHA and lime to stabilize expansive soil and studied the deformation and strength properties of stabilized soil.
Previous studies mainly focused on the physical and mechanical properties of the soil after solidification, lacking of the observation, and analysis of the microfabric during the solidification process. In this paper, the UCS experiment and the observation of the expansive soil microfabric by SEM are carried out. e changes of the mechanical properties and the microstructure characteristics are observed and analyzed during different curing time.

Basic Properties of the Test Materials.
e expansive soil used in this paper was taken from the construction site of Wanja-Yichun expressway in Yichun city, China, and the depth of the soil sample is 4-6 meters. e free expansion ratio, limit moisture content, and compaction experiment of the expansive soil were carried out under the Chinese standard-GB/T 50123-1999.
(1) e free expansion rate of expansive soil used in this experiment is 132%, which belongs to strong expansive soil (2) e plastic limit, liquid limit, and plasticity index of soil samples are 36.9, 77.6, and 40.7, respectively (3) e optimum water content of the soil sample is 28% and the maximum dry density is 1.37 g/cm 3 Undamaged RHA is taken from a biomass power generation company in Huaian, China. e RHA and lime used in the experiment are shown in Figures 1 and 2. e particle size of RHA, lime, and soil is measured by laser particle size analyzer. e data of Figure 3 are obtained. e average specific surface area of the RHA is 5910 cm 2 /g and the particle size D 50 is 7.449 um; the average specific surface area of the lime is 13880 cm 2 /g and its particle size D 50 is 3.110 um. e D 50 of soil is 20.005 um. e grain size of the RHA and the lime is very small, which can fill in the soil voids to increase the compactness. e chemical composition of raw materials is shown in Table 1. As can be seen from the table, the main components of soil are SiO 2 , Al 2 O 3 , and Fe 2 O 3 , and the main component of lime is CaO. RHA is rich in SiO 2 .

Design of Experiment Methods.
It is necessary to determine the optimum ratio of RHA and lime. e RHA and lime with different mixing ratios were mixed into standard sand to make samples. After curing for 7 d, 14 d, and 28 d, the compressive strength and breaking strength of this samples were tested. It was found that when the ratio of RHA and lime was 80 : 20, the compressive strength and breaking strength of the samples were the highest at different curing ages. So this ratio is adopted to make the soil samples.
Most scholars control the content of RHA within 0-20%, so in this paper, the amount of ash added accounts for 0%, 5%, 10%, 15%, and 20% of the dry soil weight, respectively.
In order to ensure the universality of the mesoscopic analysis in this paper, the UCS of soil samples are obtained under different curing ages. By comparing with the results of other scholars' tests, the generality of the materials used in this paper is verified and illustrated. en their microfabric is observed and analyzed, and the reasons of the improvement are obtained.

Analysis of UCS of Modified Expansive Soil
e results of compaction experiment showed that the OMC of expansive soil is 28%. However, the OMC of the mixture will be greater than 28% because of the RHA's high hydroscopicity. In this test, the moisture content of mixtures is 1 times (28%), 1.2 times (34%) and 1.4 times (40%) than the OMC of pure soil. After curing to the corresponding age in the standard maintenance room, the tests of UCS were carried out under the Chinese standard-GB/T 50123-1999. e data are shown in Figures 4-6.
Compared with the pure soil at 28% moisture content, the UCS of the mixture at 34% moisture content and 15% ash content are 2.1 times, 2.6 times, and 2.8 times of the pure soil after curing of 7 d, 14 d, and 28 d, respectively. Analogously, the UCS of 20% ash-filled soil at the corresponding curing age is 2.4 times, 3 times, and 3.2 times of pure soil, respectively. It can be obtained that soil samples form a certain strength after 7 d of curing. When the soil is cured for 14 d, the strength increase is basically completed, and the strength increases slowly during the subsequent curing. e results agree with Akshaya Kumar Sabat [13] that the curing period of RHA-lime improved soil should be longer than 7 d. e UCS of the soil sample increases with the increase of the ash content and curing age when the ash ratio is less than 20%.
rough the compaction experiment, it can be obtained that the UCS of the sample is the maximum when the moisture content of the sample is 1.2 times the OMC of pure soil. erefore, the OMC of mixtures samples is 1.2 times that of pure soil. Under the OMC, the UCS of samples with 15% ash content is 1.74 times that of the samples with 10% ash content, and UCS of the tests with 20% ash content is 1.13 times that of the samples with 15% ash content. erefore, the improvement effect of soil with 15% ash content is obviously better than that of samples with 10% ash content, but the improvement effect of samples with 20% ash content is not obvious compared with that of samples with 15% ash content. erefore, 15% ash content is chosen as the optimum mixing ratio considering the engineering cost. Because RHA and lime are mixed into soil according to the proportion of 80 : 20, when the amount of ash is 15%, the lime content is 3% of the dry weight of the soil and the RHA is 12%. e optimum lime content is consistent with that of Kumar [11] (3%) and Muntohar [12] (no more than 6%). e results of Sabat [15] study showed that 10% of RHA and 4% 2 Advances in Civil Engineering of lime content were the best and Aziz [6] considered 16% RHA as the optimal ratio. e ratio obtained in this chapter was consistent with those results. erefore, the optimum mixture ratio of RHA and lime is 12% and 3% of dry soil weight, re-spectively, in this study and the conclusion of next chapter about the microanalysis of soil has universal significance.

Microfabric Analysis of Modified Expansive Soil
e mechanical and engineering properties of expansive soil mixed with RHA and lime have been studied by many researches, but there is a lack of analysis of microfabric change of soil samples during the curing period. In order to analyze the improvement mechanism of RHA and lime further, the microfabric of expansive soil was studied by SEM. e mixture (RHA and lime in the proportion of 80 : 20, ash content at 15% of dry weight of soil) at different ages (7 d, 14 d, and 28 d) as observed by SEM, and the pore structure and microscopic characteristics were obtained. e structure and development trend of hydration products of soil samples during curing and hardening are analyzed.
Because there were too many soil samples of different curing age and different amount of ash, only 15% SEM images are selected for comparative analysis.
It can be seen from Figure 7 that there is no bond between sample particles without curing, and the particles are       spaced loosely. As can be seen from Figure 8, the boundary between particles has begun to blur after 7 d of soil maintenance, and, in the square frame of Figure 8, soil particles have been bonded as a whole. In Figure 9, most of the soil samples are bonded as a whole after 14 d of maintenance, with only a few loose particles existing. From Figure 10, after 28 d of curing, the vast majority of the soil is bonded into blocks and needle-like objects can be clearly seen in the square. e cluster and flocculation materials formed by the reaction of RHA, lime, and soil were enlarged to 10000 times in the elliptical frame of the image, and the morphology of flocculating material could be observed in detail in Figure 11. e pore between particles and the morphology of cluster materials can be seen at 10000 times after curing.
As shown in Figures 12-15, when magnified by 10000 times, it is clear that there is no connection between the particles at 0 d; after 7 d, there are two needles between particles; at 14 d of curing, cluster of substances can be seen and it continues to evolve into flocculating substances after 28 d. As shown in Figure 16, because the RHA has rich pore structure

Advances in Civil Engineering
and is rich in active silica, there will be a pozzolanic reaction in the soil mixed with RHA and lime, which will produce a coagulant to fill the soil pores to increase the strength.
Ali [16] studied the soil modified by RHA and lime. e results showed that calcium silicate octahydrate, calcium aluminate trihydrate, and calcium silicate monohydrate were gradually formed in the soil. But generally speaking, the content of silica in RHA is more than 70%, so the author thinks that the aluminum element in the conclusion of Ali mainly comes from expansive soil, including montmorillonite, illite, and kaolin. erefore, the direct reaction caused by the addition of RHA and lime is the formation of calcium silicate hydrate, which gradually precipitates into crystals during the curing process, resulting in clusters and flocculation structure, as shown in Figures 14 and 15.
Since fly ash also has pozzolanic activity, the mechanism of improving soil by fly ash has been studied to some extent. It is considered that the main reasons for the improvement    Advances in Civil Engineering of soil by lime and fly ash include the following four aspects: ion exchange, carbonation, pozzolanic action, and crystallization. rough the understanding of the composition of RHA and the observation by SEM, it can be concluded that the RHA and lime will react with the soil and the product will crystallize with time. Because the RHA is an ideal pozzolanic material which can react with lime to cause geopolymerization, the active SiO 2 in RHA dissolved by Ca(OH) 2 in lime can generate the CaSiO 3 ·nH 2 O, which can fill the voids in soil. What is more, the Al 2 O 3 in soil also can react with lime and CaSiO 3 ·nH 2 O. As a result, the ettringite is formed which can be seen in Figures 12 and 16 clearly.
Besides, the addition of small particle size, such as RHA and   Advances in Civil Engineering 7 lime, will lead to agglomeration of soil. Under these effects, the porosity of soil will decrease for the long term.

Conclusion
e following conclusions can be drawn: (1) Because of the high hydroscopicity of RHA, the UCS reaches its peak when the OMC of mixture is 1.2 times than pure soil. (2) Considering the strength and the engineering cost, 15% ash content is the optimum mixing ratio, in which the ratio of RHA to lime is 80 : 20. (3) e curing age has a significant effect on strength.
e soil sample has a certain strength at 7 d of curing, and the strength increase is basically completed after 14 d. (4) e acicular fibers produced in the pores of the modified expansive soils increased gradually during the curing period and finally formed flocculating structures. ese fibers are formed by active silica in RHA and calcium hydroxide in lime. e RHA has high pozzolanic activity and this progress is geopolymerization. (5) e main reason for the strength increase of the modified soil is that the crystal fills the pores of the soil so that the soil particles are connected together. On the other hand, the addition of RHA and lime has a certain filling effect because of their smaller particle size which plays a certain role in the solidification of soil.

Data Availability
e numerical data used to support the findings of this study are included within the article. All the lab test data and calculation results data used to support the findings of this study are also available.

Conflicts of Interest
e authors declare that they have no conflicts of interest.