Study on the Compaction Effect Factors of Lime-treated Loess Highway Embankments

This paper presents a study to investigate the effects of water content, lime content and compaction energy on the compaction characteristics of lime-treated loess highway embankments. Laboratory compaction tests were conducted to determine the maximum dry density ρdmax and optimum water content wopt of loess with different lime


Introduction
Loess is one of the wind deposited soils, which is widely distributed and constitute about 10% of the total land area of the world [1][2][3].China has a large area of loess soil deposits in the world (about 6.3×105  2 ).The world famous Loess Plateau is located at northwest of China, which occupies more than 6% of the territory of China [4][5][6] (Figure 1).At natural state, loess has high strength and small deformation.Therefore, loess is widely used as fill materials in the construction of embankments in Loess Plateau area.However, loess is regarded as one of the worst problematic soils due to it's special properties of water sensitivity.Once loess is wetted, with the effect of geostatic stress and additional pressure, the internal structure of loess is collapsed.This phenomena is defined as water collapsibility [6][7][8][9][10].Because of water collapsibility of loess, it often leads to the non-uniform settlement of loess embankments which responsible for many highway hazards in Loess Plateau area (Figure 2).In order to improve the workability and mechanical behavior of loess, lime is often added to loess as mixed fill materials of loess embankments.Lime treatment is an efficient way to improve the shear strength, plasticity and mechanical behaviors of loess through a series of physical-chemical reactions, including hydration, cation exchanges and pozzolanic reaction [11][12][13][14][15][16].When lime is added to loess, lime induces the reduction of water content through hydration and produces large amount of.Then, it leads to flocculation due to the cation exchanges of which contributes to the increase of loess plasticity and workability.Furthermore, pozzolanic reactions induce the formation of cementitious compounds which lead to the increase of shear strength and stiffness of loess [17][18][19][20][21][22].
In the construction of loess embankments, compaction is used as an efficient technique to increase the bearing capacity and decrease the non-uniform settlement and permeability of embankments.Compaction is a mechanical process by which mass of soil, consisting of solid loess particles, air and water, is reduced in volume by the application of loads [23][24][25][26][27][28][29][30][31][32].The process of compaction increases the shear strength while decreasing the compressibility and permeability of soil.Compaction quality is largely depending on chemical composition, particle size, water content of soil and compaction method.
In the past few decades, soil compaction has received increasing attention in the geotechnical engineering.Lambe (1958) investigated the effects of different compaction conditions on the mechanical properties, such as strength, permeability and stress-strain modulus of Boston blue clay.The research results were useful for interpreting the effect of compaction condition on the mechanical properties of clay [35].Simth and Dickson (1990) investigated the effects of vehicle weights and ground pressure on soil compaction by a series of field compaction tests.It was reported that an increase in wheel load produces a significantly increase in compaction quality [36].Lawton, Delage and Watabe (1991, 1996and 2000) researched the mechanical performances of compacted embankments influenced by shear strength of soils.It was shown that the shear strength of test soil has contributed to improve the compaction quality of compacted embankments [37][38][39].Horpibulsuk (2008) proposed a model to describe the compaction curves of fine-grained and coarse-grained soils under different compaction energy.These curves were useful for quick determination compaction curves by using a single trial test result [40].Jotisankasa (2009) clarified the influence of suction on the mechanical behavior of compacted silty clay by laboratory tests [41].Patel and Mani (2011) performed a field compaction tests on sandy loam soil to investigate the subsoil compaction characteristics at ranged wheel loads and multiple passes in terms of dry density and penetration resistance [42].Shouji (2014) presented the relationship between compaction condition and mechanical properties of saturated specimens through a series of laboratory tests [43].These studies investigated the compaction effect factors of many kinds of soils (clay, fine-grained soils, coarse-grained soils and sandy soils) under different compaction conditions.However, the effects of water content, lime content and compaction energy on the compaction characteristics of lime-treated loess are not understood well.
In this study, a series of laboratory compaction tests and in situ tests were performed to research the compaction influence factors of lime-treated loess embankments.Laboratory compaction tests were taken to determine the effects of water content, lime content and compaction energy on the maximum dry density and optimum water content of different lime-treated loess.In situ tests were performed to obtain the in situ dry density and degree of compaction, and to study the in situ compaction characteristics of different lime-treated loess.Optimum water contents and maximum dry density of different lime-treated loess were obtained by laboratory compaction tests.The relation curves of water content and degree of compaction were obtained by in situ tests.The effects of lime content and compaction energy on the optimum water content and maximum dry density were discussed.A comparison between laboratory compaction tests and in situ tests were made to determine the effects of in situ water contents on the compaction quality of field loess embankments.Such studies would provide valuable information for the construction of loess embankments in Loess Plateau area.

Characteristics of Test Loess
The test loess was collected from borrow pit 1# which located at Wuding highway (Figure 3).The test loess has 89.3% of the particles are smaller than 0.075  and 10.7% of the particles larger than 0.075  (0.075-2 ).The liquid limit, plasticity limit and Plasticity index, are 28, 15 and 13%, respectively.According to Test Methods of Soils for Highway Engineering (JTG E40-2007) [33], the test loess can be classified as low liquid limit clay.The physical properties of test loess are shown in Table 1.

Samples preparation
First natural loess should be air-dried and then pass through the sieve with 20  opening size.In order to avoid the influence of large particles on the compaction quality, all air-dried particles should be smaller than 20 .After drying and sieving procedure, lime was mixed with air-dried loess to obtain the predetermined lime contents (0, 3, 5 and 8%).Then water was added to each type of lime-loess mixture to reach the different target water contents (range from 8 to 16%).Finally, the obtained lime-treated loess samples were preserved at a moist container for at least 48 hours for water homogenization.Therefore, lime-treated loess samples with different water contents (range from 8 to 16%) and lime contents (0, 3, 5 and 8%) were obtained for compaction tests.

Laboratory test procedures
According to the Test Methods of Soils for Highway Engineering (JTG E40-2007) [33], multifunctional electric hammer apparatus (TG-007) which has a 4.5  rammer and a cylindrical metal mold with 100  internal diameter and 127  height was used as laboratory compaction tool to determine the relation curves between water content and dry density of different lime-treated loess (Figure 4).During the compaction tests, loess samples prepared for compaction were placed in the cylindrical metal mold with the hammer drop from a height of 450 .The detailed parameters of multifunctional electric hammer apparatus are shown in Table 2.

. Parameters of laboratory compaction apparatus
In order to determine the effects of water content, lime content and compaction energy on the compaction characteristics of different lime-treated loess, two compaction procedures were selected in laboratory compaction tests.
The first procedure was the standard compaction test (JTG E40-2007) [33].Different lime-treated loess samples at selected water contents and lime contents were placed in the mold in five layers and each layer was compacted by 27 blows of the rammer, with a total compaction energy of 2687 / 3 .The test results of the standard compaction tests are shown in Figure 8 and Table 5.
The second procedure was the different compaction energy test.Lime-treated loess samples used for this compaction tests were the same as that used for standard compaction tests.All samples were compacted using the same apparatus to produce five different compaction energies of 1990, 2687, 3383, 4080 and 4777 / 3 respectively.The number of blows of every layer were 20, 27, 34, 41 and 48 respectively.The results of different compaction energy tests are shown in Figure 9 and 10.

Test Site
The in situ test site was located at Wuding highway which cross the largest loess area of Shaanxi province (Figure 5).Loess collected from borrow pit 1# was mixed with lime, and be used as fill materials in the construction of this experimental embankment.Loess properties were the same as utilized in laboratory compaction tests.Table 3 presents the design parameters of fill materials in the construction of Wuding highway.

In Situ Test Procedures
The experimental embankment was divided into 4 parts (100 m length and 24.5 m width of each part).Each part was divided into 5 layers of h = 25 cm in loose thickness each, which was compacted with the same compaction procedure.Loess with lime content of 0, 3, 5 and 8% were used as fill materials for different parts of this experimental embankments.In order to achieve optimum compaction, water content of each layer was determined by laboratory compaction tests (range from 8 to 16%).During in situ tests, lime was added to loess by using soil stabilizing mixer.Water was added to lime-treated loess by using sprinkling truck (Figure 6).After water and lime conditioning, static roller was used for the initial compaction and final compaction.Sheep-foot roller and vibratory roller were used for re-rolling (Figure 7).The compaction procedure of each layer is shown in Tab 4.The roller combination and the actual pass times of rollers utilized in this in situ tests, were determined by the technician experiences from other field embankment construction projects in Loess Plateau area.During compaction, the roller speed was kept relatively constant at average of 3 /ℎ.In order to achieve required level of compaction for each layer, a total of 8 compactor passes were performed.

Results and Discussion
Through the standard compaction tests, the compaction curves of different lime-treated loess were achieved (Figure 8).The relation curves of compaction energy and the compaction parameters (optimum water content and maximum dry density) were obtained by different compaction energy tests (Figure 9 and 10).The relation curves of in situ water content and degree of compaction were determined by in situ tests (Figure 12).In the following, the effects of water content, lime content and compaction energy on the compaction parameters (optimum water content and maximum dry density) of different lime-treated loess will be researched.The degree of compaction of field loess embankments and the comparison of compaction curves obtained from laboratory tests and in situ tests will be discussed.Figure 8 shows the compaction curves of dry density   and water content  of different lime-treated loess under the total compaction energy level of 2687 / 3 .As it can be seen, for different lime-treated loess samples, with the increase of water content , the dry density   of each sample increased until it reached it's maximum value.The maximum value is defined as the maximum dry density   and the corresponding water content is defined as optimum water content   .As it can be seen, for all different lime-treated loess, dry density   decreases once water content  exceeds optimum water content   .

Effect of Lime Content
The effect of different lime contents on the compaction characteristics of different lime-treated loess are also illustrated in Figure 8 and Table 5.It is shown that the shapes of compaction curves of different lime-treated loess samples were similar.For different lime contents of 0, 3, 5 and 8%, the optimum water content of lime-treated loess samples were 11.7, 12.2, 12.8 and 13.6% respectively.The corresponding maximum dry density of each optimum water content were 1.91, 1.795, 1.777 and 1.764, respectively.It is note that, the addition of lime induces the increase of optimum water content   and the decrease of maximum dry density   .When lime is added to loess, water content reduces due to the process of hydration and evaporation.In addition to this process also increases the flocculation of loess particles and produces cementitious compounds which leads to the decrease of loess plasticity.Therefore, the mixed particles become more difficult to be compacted.With the increase of lime content, more water will be needed to obtain the optimum compaction.Figure 9 presents the relation curves of compaction energy and optimum water content   under the different compaction energy.Figure 10 presents the compaction curves of compaction energy and maximum dry density   under the different compaction energy.As it is shown in Figure 9 and 10, For 0, 3, 5 and 8% lime-treated loess, with the increase of compaction energy (from 1990 / 3 to 4777 / 3 ), the optimum water content   decreased and the corresponding maximum dry density   increased.It is important to note that, for different lime-treated loess, a higher compaction energy results in a higher value of maximum dry density   and a lower value of optimum water content   .

Degree of Compaction
During the construction of multi-layer highway embankments, the degree of compaction  is an important parameter which used to control the compaction quality of soil embankments.Only when the designed  value of previous layer have been achieved, the successive layer should be allowed to be compacted.The degree of compaction  can be calculated by: Where  is degree of compaction (%);  − = in situ dry density (g/cm 3 );   =maximum dry density (g/cm 3 );  = natural density (g/cm 3 );  − = in situ water content (%); The value of   is obtained from laboratory compaction test.The value of  is obtained from in situ tests by using sand cone method (Figure 11).The test results are shown in Figure 12   Figure 12 shows the relation curves of water content  and degree of compaction  for different lime-treated loess under the same in situ compaction procedure.As it shown in Figure 12, for different lime-treated loess, the relation curves of water content and degree of compaction  exhibit the same shape.With an increase of water content, the value of  increased until it reach the maximum.For 0, 3, 5 and 8% lime-treated loess, the maximum value of  were 98.9, 97.1, 96.8 and 95.8%, respectively.The corresponding optimum water content were 13.8, 12.8, 13.6 and 15.2%, respectively.It can be seen that, the value of  can hardly be achieved 100% in the practical construction of embankments.The increase of lime content induces the decrease of degree compaction  and increases the optimum water content   .In order to achieve the optimum compaction quality of lime-treated loess embankments, higher water content and compaction energy is needed during the practical construction of loess embankments.13, compaction curves from the laboratory and in situ compaction tests at different compaction conditions exhibit nearly the same shapes.For 0, 3, 5 and 8% lime-treated loess,   obtained from laboratory compaction tests was 1.91, 1.795, 1.777 and 1.764, respectively.The corresponding   was 11.7, 12.2, 12.8 and 13.6%, respectively.Based on the results of in situ tests, for 0, 3, 5, 8% lime-treated loess, the in situ value of  − was1.89,1.743, 1.72 and 1.69, respectively.The corresponding  − was 13.8, 12.8, 13.6 and 15.2%, respectively.
As it can be seen, the in situ value of  − was lower than   obtained from laboratory tests.However, the corresponding in situ value of   was greater than   obtained from laboratory tests.It is also important to note that, the situ value of   achieves it's maximum value when in situ water content was larger than the value of optimum water content   (+1-2%).During the construction of loess embankments, in order to reach optimum compaction, it is important to keep the value of in situ water content  − is larger than the value of optimum water content   (+1-2%).

Conclusion
 For different lime-treated loess, with an increase of water content , dry density   increases until it reach the maximum value   .Once water content exceeds optimum water content   , dry density   decreases dramatically.
 The addition of lime induces the increase optimum water content   and the decrease of maximum dry density   .With the increase of lime content, more water will be needed to obtain the optimum compaction of limetreated loess.
 With an increase of compaction energy, the maximum dry density   decreases and the optimum water content   increases.In order to achieve optimum compaction, it is useful to increase compaction energy.
 The value of  can hardly be achieved to 100% in the field construction of embankments.The increase of lime content induces the decrease of degree compaction  and increase the optimum water content   .
 The situ value of  − achieves it's maximum value when in situ water content was larger than the value of optimum water content   (+1-2%).In order to achieve the optimum compaction quality of lime-treated loess embankments, higher water content and compaction energy is needed during the practical construction of loess embankments.

Figure 1 .
Figure 1.Loess Plateau area located at northwest of China.(a) Loess area at Yan'an city; (b) Loess area at Yulin city

Figure 2 .
Figure 2. Highway hazards in Loess Plateau area.(a) Embankment hazards of Tianding highway; (b) Pavement hazards of Yanyan highway

Figure 6 .
Figure 6.(a) Soil stabilizing mixer; (b) Sprinkling truck of compaction for each layer, a series of in situ tests were performed throughout the test area.The tests were included water content tests and natural density tests.Each of these tests were conducted in accordance with Field Test Methods of Subgrade and Pavement for Highway Engineering [JTG E60-2008][34].

Figure 8 .
Figure 8. Compaction curves of different lime-treated loess samples

Figure 9 .Figure 10 .
Figure 9. Relation curves of compaction energy and optimum water content and 13.