Mechanical Behaviors of Natural Sand Soils andModified Soils in Heavy-Haul Railway Embankment

'e addition of chemical or mechanical materials, such as fibers or stabilizers, is frequently utilized in geotechnical engineering to improve the mechanical properties of problematic soils. In this study, great efforts have been made to obtain insight into the mechanical properties of the natural, fiber, and chemical additive-stabilized soil in heavy-haul railway embankment. A series of triaxial compression tests are conducted on the stabilized samples of different preparation conditions, including water content, compaction degree, confining pressure, fiber content, fiber length, stabilizer content, and curing time. Results show that the shear strength of natural soils shows a distinct increase after adding fiber and chemical additive stabilization. 'e optimum fiber content and length for fiber stabilization are 0.2% and 12mm, respectively. 'e initial tangential modulus and failure stress of chemical stabilized samples increase with the increase of additive dosage or curing time. Meanwhile, a brittle characteristic is observed. In the process of determining the reinforcement methods in practical projects, several other considerations are included, such as equipment and time available, especially for stabilized soils. 'e fiber-reinforced soils and stabilized soils are efficient for increasing the shear strength and changing of the brittleness character of the heavy-haul railway embankment. 'e results of this study could provide a valuable reference for geotechnical engineers dealing with soil problems, especially for the heavy-haul railway embankment.


Introduction
In practical projects, the diseases of heavy-haul railway include settlement, bed deformation, and shoulder extrusion. e thesis is based on the Bazhun heavy-haul railway project, located in the north of China. Investigations were conducted to identify the dynamic deformation module of the embankment. However, the natural soils of the embankment failed the quality of heavy-haul railway embankment that induces the engineering problems. e purpose of the treatments is to satisfy the quality and safety of the design life of the project. ere are numerous means to achieve this purpose, including replacement filling, compaction pile, geosynthetic materials reinforcement, chemical reinforcement, and other methods. Replacement filling is commonly used in the treatment of subgrade soils, but it has a high cost with the limitation of replacement material and transportation fees. Compaction pile should be used in the site with saturated clay, which is not suitable for the sandy soils. In some projects, geotextiles are selected to improve soil mechanical properties due to their highstrength, economical, anticorrosion ability. e utilization of additions including mechanical and chemical methods or combination of these two is frequently used in engineering to improve the strength and the stability of natural soils.
ere are two different types of chemical stabilizers: grouting materials and powder materials. For the former one, it includes cement, clay, water glass, lignin, propylene, epoxy resin, and urinary resin. For the late one, it includes cement, quicklime, fly ash, gypsum powder, and many other relative materials. However, most of the chemical stabilizers could induce environmental problems to some extent and, thus, it is necessary to develop a new environment-friendly material to stabilize the natural soils.
Soil stabilization attracted considerable interest among the above methods. e utilization of additions including mechanical and chemical methods, or combination of these two, is frequently used in engineering to improve the strength and the stability of natural soils. Many practicing projects adopted this method, such as repairing failed slopes [1], Earth retaining structures [2], antiearthquake geostructures [3], railway embankment [4][5][6], airfield construction, and other geotechnical engineering applications.
e previous literature indicates that both fiber reinforcement and stabilizer reinforcement improve the shear strength and ductility of soils. A large number of direct shear tests [7][8][9], triaxial compression tests [10][11][12][13], and field tests [14,15] have demonstrated that the shear strength has a large increase with the fiber reinforcing considering different influence factors such as confining pressure, fiber content, fiber orientation, and fiber properties. Fibers interact with the soil particles when bearing forces through surface friction and interlocking. e function will help to transfer the stress from the soil to the fiber. Besides, the tensile strength of fibers is mobilized and imparted to the soils resulting in the decrease of strain and the improvement of soil engineering properties. Another method employed by projects frequently is stabilizer reinforcement, which can improve the soil strength and increase resistance to softening.
ere are many cementitious materials utilized to stabilize the soils such as cement, lime, fly ash, bitumen, or combination of these [16][17][18][19][20][21][22][23]. e variation of mechanical properties of stabilized soils is closely related to the presence of organic matters, sulphates, sulphides, and carbon dioxide. e compaction condition, moisture content, temperature pozzolanic, freeze-thaw, and dry-wet cycles are also essential influence factors to the strength of stabilized soils [24]. However, these stabilizers are not suitable for many projects for non-environment-friendly and brittleness character [25].
ese previous studies indicate that both fibers and stabilizers can improve the shear strength and ductility of soils. e mechanical properties of the reinforced soils are influenced by the natural soils characters, fibers, and stabilizers properties.
In order to develop a suitable soil treatment for the unsuitable heavy-haul railway embankment, an environment-friendly new stabilizer was developed to reinforce the locally available soils. Forty-one conventional triaxial compression tests are carried out on the natural soils, fiberreinforced soils, and stabilized soils with a newly developed stabilizer, considering different water content, compaction degree, confining pressure, fiber content, fiber length, stabilizer content, and curing time. e principal object of these tests was to study the properties of the reinforced soil under different conditions. A second objective was to determine if there is optimal value, such as fiber content, fiber lengths, stabilizer content, and curing time for the tested natural soils. irdly, comparing the mechanical properties of the fiber-reinforced soils and stabilized soils helps engineers making reasonable decisions. e results obtained from this research provide a valuable reference for the design and reinforcement project of the heavy-haul railway embankment in northwest district in China, as well as the other districts owning similar natural soil in the world. Besides, it can provide useful parameter reference for quality control of reinforced soils.

Materials.
e soils utilized in the tests were collected from the third section of Bazhun heavy-haul railway, which is located in the north of China. e coefficient of uniformity (Cu) and the coefficient of gradation (Cc) are 3.5 and 2.24, respectively. e curve of particle grade of subgrade soil is shown in Figure 1. e soils collected from the site were dried, crushed, and sifted to satisfy the requirement of the Chinese code for soil test of railway engineering (TB 10102-2010, 2010). Based on the compaction test of subgrade soils, the optimum water content and the maximum dry density are 8.5% and 2.06 g/cm 3 , respectively.
Two different approaches were adopted to modify the soils, including reinforcing with discrete randomly distributed fiber and chemical stabilization in order to achieve the desired engineering properties. Polypropylene fibers (MP-I, produced by Bonnyfibres in Shanghai) were selected as reinforcement stabilizers because of the acid, alkali, and salt resistance properties with excellent dispersibility and safety. e properties of the fibers are given in Table 1.
e new proposed stabilizer was designed by Harbin Institute of Technology, which was a mixture of several natural mineral materials and inorganic substances. e component proportion ratio for the mixture is 14% of silicate, 36% of aluminate, 21% of tetracalcium aluminoferrite, 9% of sulfate, 3% of silicon dioxide, and 17% of sulphoaluminate. Besides, the colour of the mixed powder is white with a fineness over 200 μm, and it has the characters of being odorless, nontoxic, and soluble in water.

Samples of Natural Soils.
Complying with the water content requirement in test procedures, the specific samples are prepared, given a good stir, and preserved for 24 hours, ensuring the uniform moisture of the final samples. After the preserving time, the soils are capable of making test specimens.

Samples of Fiber-Reinforced Soils.
It is crucial to ensure the dispersion uniformity randomness of fibers, which is a significant factor affecting the mechanical property of mixtures in case of twine, spillover, and segregation [26]. e mixture was stirred by hand as follows: (i) Dry the soils. (ii) e required components, including water, fibers, and dried soils, are weighted precisely. Note that the water content is 2% more than the specific amount of water in the tests, aiming to avoid evaporation. hours.

Samples of Stabilized Soils.
e specific amount of stabilizer, as planned, was mixed thoroughly with dry soils and water. In order to guarantee the uniformity of the mixture, the water was added by spray successively until the water content reaches 11.5%, which is 3% more than the optimum moisture content. It was utilized in the chemical reaction process between minerals in soil with stabilizer. en the prepared material is put into the air cabinet for curing. Finally, the test specimens are prepared with specific curing time.

Test Procedures.
In this study, triaxial tests were conducted in the Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, using MTS-810 on 61 mm in diameter and 125 mm high samples. All the materials were compacted in a triaxial test mold by a temping successive layer to ensure uniform compaction. e loading rate of straincontrolled undrained triaxial consolidation experiments was 1.25 mm/min. A serviceability failure criterion was defined as 15% axial strain for compressive loading if no evident peak value was detected in the tests. Otherwise, the peak value will be taken as the failure point. A complete list of tests is given as follows, considering different factors, to analyze the mechanics of natural soils, fiber-reinforced soils, and stabilized soils.

Test Procedures of Natural Soils.
A variable-controlling approach was adopted to test the effect of water content, compaction degrees, and confining pressure levels. A summary of the tests is given in Table 2.

Test Procedures of Fiber-Reinforced Soils.
e effectiveness of the reinforcement is influenced by many factors, including fiber properties and soil properties. e effects of fiber content, fiber length, and confining pressure levels were taken into account, as shown in Table 3, with a water content of 8.5%.

Test Procedures of Stabilized Soils.
As for the stabilized soils, three factors were considered in the process of test design, which are stabilizer content, curing time, and confining pressure. Note that the stabilizer content is determined by the test output carried out by the Harbin Institute of Technology through the long-term trial-and-error approach. Based on the previous study, the best stabilizer content is about 5% to 7%; that is the reason why the 5% content was determined to testify the effects of curing times and confining pressure levels. All the tests about stabilized soils with the same water content, 11.5%, are listed in Table 4.

Typical Results and Discussion
Conventional triaxial compression tests are conducted on natural soils, fiber-reinforced soils, and stabilized soils. Typical results of the tests provided material properties for analyses.

Natural Soils Test Results.
As mentioned, natural soils were collected from the available local site. e study showed that the undrained stress-strain behavior of natural soils has a strain hardening character, so the failure criterion was defined as 15% axial strain. e effects of compaction degree, confining pressure level, and water content were analyzed in this study. Figures 2 and 3 illustrate, respectively, the deviatoric stress strain and the relationship between compaction degree and failure stress. e deviatoric stress increases with the axial strain with a turning point. After the point, the deviatoric stress increases slowly. Comparing the results of soils with different compaction degree, increasing the compaction degree generates an increase of the peak shear strength of the natural soils. e relationship between compaction degree and failure stress is almost linear, as shown in Figure 3. is linear relationship is generated because of the compaction degree, significantly affecting the void ratio. With the increase of the compaction degree, the natural soils become dense, and the addition of effective contact area between soil particles induces the improvement of interface friction. Figure 4 presents the influence of confining pressure on the stress-strain behavior of natural soils under undrained conditions. With small strain, the deviatoric stress increases quickly. After about 2% of the strain, the increase becomes gentle. e results illustrate that,     with the increment of the confining pressure, the shear strength of natural soils improved gradually. e improvement in the fail stress under different confining pressure is shown in Figure 5. e cohesion intercept and friction angle for natural soils are 0 MPa and 38°, respectively. Under confining pressure, the horizontal dilatancy was restrained, which induces the soil particles harder to slide and roll. Hence, the soil strength is improved significantly.

Effect of Water Content on the Strength Behavior of Natural Soils.
Typical results of natural soils with different water contents are presented in Figure 6. e water content heavily influences the deviatoric response of the natural soils. Figure 7 illustrates that if the water content is lower than the optimum water content, failure stress will increase with the increase of water content. However, it will decrease if the water content is increased over the optimum value. For the soils containing more than 10% water, the shear strength decreased quickly. With a low water content, the failure stress is inferior to the value of optimum water content for the disintegration of the soil particles bonding. Water content higher than the optimum value also induces the weakening of soil shear strength by the deceleration of absorption forces and friction interaction of soil particles.

Test Results of Fiber-Reinforced Soils.
e contribution of random discrete fibers on the strength of reinforced soil is remarkable. Randomly oriented fibers incorporating soil particles generated surface friction bond and interlocking effect. is function allows the stress transfer from soil to fibers, which will reduce the strains in reinforced soil, resulting in the improvement of soil strength. e principal object of this fiber-reinforced soil test was to study the properties of the fiber-reinforced soil with different fiber content, length, and confining pressure. A second objective was to determine if there are optimal fiber content and fiber lengths for the tested natural soils.

Effect of Fiber Content on the Strength Behavior of Fiber-Reinforced Soils.
Typical results of fiber-reinforced soil with different fiber contents are presented in Figure 8.
e results recorded are in line with the already published results by Heineck et al. [27]. e presence of fibers has little influence on the deviatoric stress under small strain (0.01%∼ 0.3%). As the strain developed, the effect of fiber content on the shear strength becomes apparent, especially under considerable strain (0.3%∼15%), for the surface friction bond and interlocking function contributing to the soil pressure. e failure deviatoric strength envelopes for fiberreinforced soils are shown in Figure 9, based on a serviceability failure criterion and 15% axial strain for compress loading. e fiber content heavily influences the deviatoric response of the reinforced soil and the influence has a limitation. e optimal fiber content is 0.2% for the tested soils. If the percentage of fiber is lower than it, the soils have other new voids leading to the low shear strength.

Advances in Civil Engineering
Moreover, with the increment of the fiber content, the dense degree of the reinforced soil increases gradually until the optimal point. However, the dense degree will decrease after the optimal point for the density of fiber is lower than the soil. Besides, the increment of fiber weakens the interlocking function.

Effect of Fiber Length on the Strength Behavior of Fiber-
Reinforced Soils. Soils were reinforced with various length fibers, but the same material properties were utilized to study the influence of fiber length on the modification of shear strength at confining stress of 0.1 MPa and fiber content of 0.2%. e deviatoric stress-axial strain for fiber-reinforced soil with different fiber length is illustrated in Figure 10. It was found that the longer the fiber, the more significant deviatoric stress of the soil. But the deviatoric stress decreases when the fiber length is         Advances in Civil Engineering 18 mm, over the optimal length, which is considered as 12 mm, as shown in Figure 11. is result was most probably due to the change in mechanical interaction influence. e tensile force generated in the fiber can endure the interfacial shear stress of soils and it was proportionate to the fiber length. However, for soils with short fiber, there is not enough tensile force and the surface friction bond is weak, which had little influence on the improvement of the soil strength.

Effect of Confining Pressure Level on the Strength
Behavior of Fiber-Reinforced Soils. Four different confining pressures were adopted at optimal length and optimal fiber content to identify the influence of confining pressure on the shear strength of fiber-reinforced soil.
e recorded results are presented in Figure 12. e deviatoric strength increased along with the increase of confining pressure, for higher confining pressure can enhance the interaction forces between fibers and soil particles. e failure deviatoric strength envelopes for the fiber-reinforced soils with different confining pressure are almost linear as shown in Figure 13. e cohesion intercept and friction angle for fiber-reinforced soil are 0.11 MPa and 37.7°, respectively. Also, the comparison between natural soils and fiber-reinforced soils can be generated from Figures 4 and 8. e latter has a significant improvement in the cohesion intercept but less change on the friction angle.

Test Results of Stabilized Soils.
Conventional triaxial compression tests are conducted on stabilized soils made of stabilizer and natural soils. e reported experimental study includes various stabilizer contents, curing time, and confining pressure to evaluate the feasibility of using this new stabilizer in the reinforcement of natural soils in heavy-haul railway embankment.       Figure 14, the deviatoric stressaxial strain relationship under confining pressure of 0.1 MPa and curing time of 3 days illustrates that the stabilization effect results in a prepeak restrain of the dilatancy. After the peak shear strength, the disintegration of the bonding leads the stabilized soils to undergo plastic flow, which is the different fundamental behavior between stabilized soils and natural soils. e initial stiffness, peak shear strength, and residual shearing resistance are improved because of the utilization of stabilizer. e peak deviatoric stress of stabilized soils with 5% of stabilizer is 0.67 MPa, which is 1.8 times higher than that of the natural soils. However, the postpeak stress decrease speed increase with the increase of the stabilizer content, which means that the increase of the stabilizer induced an increase in brittleness. Figure 15 illustrates the failure stress of the stabilized soils with different stabilizer content. Increasing the stabilizer content generates an increase of the failure stress. is character is because of the internal confinement or cohesion improvement when more stabilizer content is added in the soils. e internal bonding is destroyed in the extensive strain process when the shear strength is similar to natural soils. Figure 16 presents the deviatoric stress-axial strain relationship for stabilized soils with different curing times.

Effect of Curing Time on the Strength Behavior of Stabilized Soils.
ese results illustrate the effect of curing time on the shear strength of stabilized soils. e peak shear strength increases with the increase of curing time for stabilized soils, for the stabilizer undergoing a chemical reaction to a great extent when curing time is extended, as shown in Figure 17. However, the deterioration of the internal bonding generated a rapid postpeak strain softening. It is also interesting to indicate that, with the increase of the strain, the deviatoric stress tends to be gentle until the end of the tests. e comparison illustrates that, with the increase of curing time, the brittleness was aggravated.     Advances in Civil Engineering

Effect of Confining Pressure Level on the Strength
Behavior of Stabilized Soils. e response of stabilized soils with 5% of stabilizer content and 3 days of curing time to axial compressive stress under different confining pressure is illustrated in Figure 18. e results showed that, with the increase of confining pressure, the failure stress and residual shearing resistance are improved significantly. For instance, the failure stresses for the samples under 0.1 MPa and 0.3 MPa are 0.67 MPa and 1.26 MPa, respectively, with 88% increase. Besides, the residual shearing resistances for them are 0.41 MPa and 0.84 MPa, with a double increase. Figure 19 shows the result of deviatoric strength envelopes for stabilized soils, which presented the almost linear relationship between confining pressure and failure stress. Confining pressure protected the samples from horizontal dilatation, in which way the cracks inside the stabilized soils are restrained and the shear strength of the soils increased. e cohesion intercept and friction angle obtained from the test results are 0.11 MPa and 34.7°, respectively.  Figure 20. Fiber-reinforced soils in Figure 20 have a fiber content of 0.2% and stabilized soils have a stabilizer content of 5%. Evaluation of the comparison of the shear strength of these soils suggests that fiberreinforced soils and stabilized soils are significantly improved. For natural soils, there is no peak stress point in the relationship of deviatoric stress-axial strain, which is similar to the character of fiber-reinforced soils. e behavior of stabilized soil is different from those of the other two soils, showing a prepeak restrain of the dilatancy and brittleness character.
e initial tangential moduli of stabilized soil, natural soil, and fiber-reinforced soil are 134.3 MPa, 93.9 MPa, and 85.4 MPa, respectively. e initial tangential modulus of stabilized soil is higher than those of natural soil and fiber-reinforced soil. Stabilized soils obtained the peak shear strength at 2% axial strain; then the deviatoric stress decreased. Compared with stabilized soils, the other soils' deviatoric stress increased with the growing up of strain until the failure strain. e cohesion intercept and friction angle obtained from the test results of different confining pressure are given in Table 5. Note that the fiber-reinforced soils in Table 5 have a fiber content of 0.2% and the stabilized soils have a stabilizer content of 5%. e cohesion intercepts for fiber-reinforced soils and stabilized soils are increased to 0.11 MPa; at the same time, the friction angle generated a decrease. Figure 21 presents the failure stress for different soils at confining pressure of 0.1 MPa. e fiberreinforced soils in Figure 21 from left to right have fiber content of 0.1%, 0.2%, and 0.3%, respectively. e left and the right stabilized soils columns shown in Figure 21 have a    stabilizer content of 3% and 5%, respectively. ere is an increase in failure stress for both fiber-reinforced soils and stabilized soils. e failure stress for stabilized soils is higher than that for fiber-reinforced soils. Besides, with the change of fiber content and stabilizer content, soils showed different failure stress. Figures 22-24 show the appearances of soils after failure. As shown in Figure 22, natural soils generated ballooning excessive damage with no crack. For fiber-reinforced soils, the specimens illustrated a deformation in shear failure and ballooning excessive damage, as shown in Figure 23. Figure 24 contains three different stabilized soils, cured for 3 days, 14 days, and 28 days, respectively. All the stabilized soils showed a brittle fracture and a deformation in shear failure. For stabilized soils cured for 3 days, the specimens showed ballooning excessive damage without crack. For stabilized soils cured over 14 days, a visible crack appeared after failure with more than one shear fracture plane. With an increase of curing time, the stabilizer had better chemical reactions with soil particles creating much more crystallization outcomes and consuming more water. e crystallization and matric suction result in the increase of soil strength and brittleness.

Conclusions
is study was undertaken to investigate the mechanical properties of natural soils in heavy-haul railway embankment, fiber-reinforced soils, and stabilized soils by performing conventional triaxial compression tests. e results of this study bring forth the following conclusions: (1) e undrained stress-strain behavior of natural soils has a strain hardening character. Considerable increase of initial tangential modulus and failure stress    was induced by the increase of compaction degree and confining pressure. e water content heavily influences the deviatoric response of the natural soils, and, with the optimum water content, 8.5%, the soils performed better. (2) e deviatoric stress-axial strain relationship for fiber-reinforced soils has no peak shear strength point, similar to natural soils. e improvement of initial tangential modulus and failure stress of fiberreinforced soil is limited with optimum content and length, 0.2% and 12 mm, respectively. e failure stress for the fiber-reinforced soils grows linearly as confining pressure increases.
(3) Stabilized soils generated a peak deviatoric stress in the process of axial compression and the brittleness of soils increased if fiber contents increased. With the increasing of fiber contents, curing time, and confining pressure, the initial tangential modulus and failure stress for stabilized soils increased, which are superior to the fiber-reinforced soils. e residual shear strength increases with the increase of stabilizer content and confining pressure. However, it shows a decreasing character with the increasing of curing time. (4) Shear strength increases as a result of fiber reinforcement and stabilizer reinforcement for the natural soils in heavy-haul railway embankment. e reinforced soils are efficient for heavy-haul railway embankment, which is close to the ground surface.
In the process of determining the reinforcement methods, several other considerations are included, such as equipment and time available. (5) It should be pointed out that these results are valid only for the natural soils utilized in this study. However, the results can also provide guidance or references to the other similar projects. In addition, the dynamic load and scale effect on the mechanical properties of the reinforced soils have not been investigated thoroughly. Hence, further studies including dynamic stabilities and large-scale tests are needed to better understand the characteristics of fiber-reinforced soils and stabilized soils.

Data Availability
Some data used during the study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare no conflicts of interest.