Analysis of Deformation Characteristics of Filled Subgrade Treated by Micro-Pile

School of Mechanics, Civil Engineering & Architecture, Northwestern Polytechnical University, Address: No. 1 Dongxiang Road, Chang’an District, Xi’an 710129, Shaanxi, China Key Laboratory of Highway Engineering in Special Region, Ministry of Education, Chang’an University, Address: No. 126, Middle Section of South Second Ring Road, Beilin District, Xi’an 710064, Shaanxi, China College of Geology and Environment, Xi’an University of Science and Technology, Address: 58 Yanta Middle Road, Beilin District, Xi’an 710054, Shaanxi, China


Introduction
A micro-anti-slide pile-also known as a "micro-pile"-is a small anti-slide pile with a diameter of less than 300 mm that can be constructed by drilling a hole, reinforcing the structure, and filling the hole by pressure grouting. As micro-piles are easy and quick to construct and can be flexibly positioned according to the needs of the project, they are widely used in slope engineering projects for reinforcement and slip prevention [1][2][3][4].
ough micro-piles were initially used for foundation reinforcement, they have more recently been applied to a variety of advanced engineering projects including landslide treatment and deep foundation pit support [5][6][7]. In slope reinforcement projects, the ability to construct micro-piles rapidly is particularly valuable in emergency projects, such as those for landslide treatment; furthermore, micro-piles offer an advantage over traditional anti-slide piles as they can be constructed with less disturbance to the slope. Similar to traditional anti-slide piles, micro-piles are under the same force as the rock and soil body following construction and work to maximize the anti-sliding ability of the rock and soil body [8][9][10].
A variety of previous studies have investigated the applications of micro-piles. Wu et al. [11] studied a combined support system comprised of micro-piles and a concrete sprayed layer, which maximized the uplift capacity of the micro-piles and improved the traditional soilnailed wall support form. Fang et al. [12] explored the antisliding effect of coupling beam single-row anti-slide piles through outdoor model tests. e tests showed that coupling beams can reduce the top displacement and pile body damage of micro-piles as well as the shear strength of the sliding surface. Liang et al. [13] investigated two types of pile core and pile circumference reinforcement and found that the moment of inertia of the pile section under the circumference reinforcement is greater than that of the pile core. Zhu et al. [14] conducted a study on the use of miniature anti-slide piles to strengthen the ultimate resistance of a slope and found that miniature anti-slide piles can bear the bending and shearing action for the slope as a whole. Zhu et al. [15] found through experiments that the horizontal bearing capacity of a micro-pile is directly related to the type of slope rock and soil in which it is embedded in addition to the conventionally considered parameters-pile section, flexural rigidity, and pile layout. Esmaeili et al. [16] introduced a successful case of using micro-piles to reinforce the slope of a road embankment in the Commonwealth of Pennsylvania, USA. Chen et al. [17] used numerical simulation methods to study the influence of geotechnical parameters on the bearing capacity of micro-piles and found that the friction angle has the greatest influence on the horizontal bearing capacity of micro-piles.
In summary, there is a clear relationship between the force mechanism of a micro-pile and the pile dimensions, the surrounding rock and soil type, and the nature of the load on the pile. However, evaluating the deformation of a subgrade after micro-pile treatment remains challenging.
is study aims to address this challenge by researching the micro-pile treatment of the filled subgrade along the K555 + 070 to K557 + 710 section of the Qingdao-Yinchuan Expressway. To this end, the deformation and stress of the subgrade were monitored and Midas/GTS software was used to create and validate a numerical simulation of this project based on the monitoring data.

Study Area and Monitoring Setup
2.1. Study Area. Section K555 + 070 to K55 + 710 of the Qingdao-Yinchuan Expressway is located in Ningjin County, Hebei Province. e study area is located in the central and southern region of the Jizhong Plain, which is an alluvial plain at the eastern foot of the Taihang Mountains. e terrain is low and flat, sloping slightly from northwest to southeast with a natural slope of 1/4000. e highest elevation in the whole region is 36.5 m above sea level, and the lowest elevation is 24.4 m above sea level. e average annual precipitation is 449.1 mm, with large seasonal and interannual differences in precipitation. ere are 11 seasonal rivers in the area. e total length of the river is 170.32 km, the density of the river network is 0.16 km/km 2 , and the total annual runoff is 3.154 billion cubic meters.
On-site drilling and indoor geotechnical tests of the section K555 + 070 to K557 + 710 show that the soil in this section is mainly composed of silt, silty clay, and clay, with a layer of medium-fine sand interspersed locally. e cracks develop along the direction of the roadbed (see Figure 1), and the cracks are 0.5-15 mm wide. From the appearance of the crack to the micro-pile treatment, the crack continued to expand, and secondary cracks were generated at the cracks, especially in the rainy season. e road maintenance work area once used the cracked pavement to be filled with joints, but the treatment effect was not obvious.
In order to control the continuation of the uneven settlement of the roadbed, it is planned to carry out grouting + root pile reinforcement treatment on the road section. After the construction is completed, the micropile, the grout around the pile, and the original roadbed work together.
e strength of the micro-pile body is ≥20 MPa, and the bearing capacity of a single pile is ≥150 KN.
According to the filling height of the roadbed, the micro-pile design row spacing is 3 m, the row spacing is 2.5-4.0 m, the pile length is 9.5 m, and the pile diameter is 140 mm. According to field tests, the single-hole grouting radius of the micro-pile is about 1.5 m. e grouting filling factor is 3 to 5%, and the grouting amount per unit length is 0.1875 to 0.4875 m 3 . e ratio of micro-pile diameter to pile length in this project was small. us, under the road load, the pile top will have some perforation effect on the road surface. However, the grouting effect resulting from the micro-pile construction process significantly improves the strength of the soil and the load-bearing capacity of the roadbed.

Pressure Monitoring.
We implemented a physical test project; to this end, 36 soil pressure boxes were buried in the soil near the 12 micro-pile bodies at the K557 + 348 culverts, and 6 soil pressure boxes were buried in the soil near the 2 micro-pile bodies at K557 + 680 ( Figure 2).

Deformation Monitoring.
After the completion of the roadbed construction on August 7, 2011, the K557 + 000 to K557 + 680 section was subdivided into seven cross sections with six observation points each, for a total of 42 observation points. ese observation points were used to evaluate the settlement and deformation characteristics of the subgrade. e reference point was located on concrete that is not affected by settlement. It is measured by Leica NA2 automatic level and micrometer with an accuracy of 0.3 mm/km. We found that the settlement of the subgrade center during the operating period of the roadway and after the micro-pile treatment was 25.60 cm and 0.55 cm, respectively. e horizontal displacement at the toe of the slope was 4.01 cm during roadway operation and 0.04 cm after micro-pile treatment. e horizontal displacement at the shoulder was 2.72 cm during roadway operation and 0.08 cm after micropile treatment. 2 Advances in Civil Engineering

Simulation Assumptions and Regions.
We used Midas/ GTS NX software for modeling. Taking into account the complex geological environment as well as the actual filling and treatment of the roadbed in the study area, the following assumptions were made in developing the simulation model: (1) e rock-soil mass is an ideal elastoplastic body and obeys the Mohr-Coulomb strength yield criterion. (2) e influence of groundwater and seepage on the whole project is simplified. (3) e micro-pile is a completely elastic body, and the vehicle load during roadway operation can be simplified to a static load. (4) e rock-soil-structure interface is defined by the pile contact characteristics. e grouting effect refers to the increase in the strength of the soil around the pile.
After careful consideration, we designed the model to represent a 1 m section along the roadway with a crosssectional foundation width of 78 m and a foundation thickness of 30 m. e height of the subgrade filling was 6.6 m, the top surface of the subgrade was 28.0 m wide, and the bottom surface was 46.0 m wide. e micro-piles were spaced 1.625 m, 2.50 m, 3.75 m, and 4.00 m from the center of the road. e foundation consisted of five layers of soil. From top to bottom, they were the hard shell layer, clay layer, silt layer (including sand), silt soil layer, and silty clay layer, which were 0.5, 2.2, 3.2, 4.9, and 4.0 m thick, respectively. e subgrade fill was divided into five layers with thicknesses of 0.5, 2.5, 1.5, 1.3, and 0.8 m from top to bottom. A schematic of the model is shown in Figure 3.

Model Boundary Conditions.
e model boundaries and constraints were as follows: e initial stress used for model calculations was the selfweight stress of the roadbed. e model was divided into 10552 units and 10758 nodes.

Calculation Process.
We divided our analysis into three stages as follows: completion of roadbed construction (6 months), roadway operation period (24 months), and postmicro-pile treatment (12 months). An overview of the calculation process is shown in Figure 4.

Physical and Mechanical Parameters.
e parameters used for the micro-piles were determined according to the construction specifications. Poisson's ratio and unit weight were determined from geotechnical tests. e elastic modulus, friction angle, and cohesion were determined based on geotechnical test data and adjusted based on the deviation rate determined after performing trial calculations. Tables 1 and 2 summarize the model parameters.

Validation of Calculated Results.
e simulation calculations for the existing roadbed are largely dictated by several key geotechnical parameters. Notably, strategies for the selection and determination of model parameters have always been controversial. e current parameter selection methods use data obtained through geotechnical tests, statistical data based on similar strata, or empirical data. However, there are limitations associated with each of these approaches. Since the parameters determined by geotechnical tests differ from the parameters of actual projects, they need to be corrected. e statistical data method is only applicable to ordinary stratum and requires the accumulated knowledge from many similar engineering projects. Finally, empirical data are convenient to use but lack scientific rigor. In short, none of the three existing parameter selection methods can be applied universally, which is a significant limitation for their use in special geological environments.
us, identifying a method that can use on-site monitoring data to validate the main parameters for theoretical calculations and finite element analysis is very important. We propose validating the simulation parameters by calculating the deviation rate (P). When the deviation rate between the simulation result and the monitored value is within the allowable range (P < 50%), the physical and mechanical parameters selected by the calculation model can be considered reasonable.
In our study area, on-site monitoring revealed that the settlement of the subgrade center during the roadway operation period and after the micro-pile treatment was 25.60 cm and 0.55 cm, respectively. e simulation results show that the vertical displacement of the embankment center during the roadway operation period and after the micro-pile treatment was 21.43 cm and 0.35 cm, respectively. We calculated the deviation rate as (calculated value−monitoring value)/monitoring value and found that the simulated data deviation rate during the operation period was −16.30% < 50%; furthermore, the simulated data deviation rate after the micro-pile treatment was −36.36% < 50%. us, the parameters selected for the simulation are valid.

Results
e deformation in the direction perpendicular to the ground plane is the most important parameter of uneven settlement and destruction in the process of subgrade filling, operation, and treatment. At the same time, considering the need to use the subgrade settlement data obtained from the previous deformation monitoring to check the simulation results, the Y-direction displacement simulation result analysis is selected here.
Taking the centerline of the roadbed as the axis of symmetry, it can be seen in Figures 5-7 that the underside of the side slope on both sides of the roadbed, the side slope line, and the road shoulder are the most concentrated areas of deformation. e settlement pattern of the subgrade center is arcshaped and slopes down to the surroundings, whereas the shoulders on both sides of the center axis shift towards the center of the subgrade. In the three considered working conditions-roadbed construction, roadway operation, and micro-pile treatment-the maximum horizontal displacement directly below the toe was 4.05 cm, 1.37 cm, and 0.03 cm, respectively. e maximum horizontal displacement at the toe was 1.61 cm, 3.81 cm, and 0.03 cm, respectively. Finally, the maximum horizontal displacement at the shoulder was 6.54 cm, 2.53 cm, and 0.06 cm, respectively.
No micro-piles were added during the operation period, and a large portion of the foundation soil was compacted by roadbed filling and vehicle loads. During this period, the horizontal displacement of the center of the embankment was zero, and the maximum vertical displacement was 21.43 cm. However, over time, both the horizontal and vertical displacements at the toe of the embankment slope tended to increase.     Figure 7: Model results for vertical displacement following micro-pile treatment. Table 3 summarizes the simulation calculation results which show that the deformation of the road shoulder and the center of the roadbed has the law of roadbed construction > roadway operation period > post-micro-pile treatment. It shows that the compression and consolidation of the fill account for most of the settlement and deformation of the roadbed in the whole life cycle. Both the monitoring data and simulation calculations show that the maximum settlement of the roadbed center is far greater than the maximum deformation of the slope toe and the shoulder in the X-direction. is shows that it is critical to consider the settlement of the center of the subgrade when monitoring subgrade deformation. Both the monitoring data and simulation calculations also show that the maximum deformation of the slope toe, the shoulder, and the center of the roadbed decreases by more than 96% after adding micro-piles. Specifically, the settlement of the roadbed center decreases by more than 98% after micro-pile treatment, indicating that the micropile has a significant effect on mitigating the uneven settlement of the roadbed.

Conclusion
is article relies on the micro-pile to treat the uneven settlement of the roadbed. We use on-site monitoring and numerical simulation methods and analyze the stress and deformation characteristics of the subgrade before and after the micro-pile treatment. e conclusions are drawn as follows: (i) We propose and use the deviation rate P to check the new method of simulation results. e compression and consolidation of the filled soil account for most of the settlement and deformation of the roadbed in the whole life cycle.
(ii) Comparative analysis of monitoring and simulation results shows that the settlement of the center of the roadbed is a sensitive part of deformation monitoring. After piles are added, the deformation extremes at the toe, shoulder, and center of the roadbed are all reduced by more than 96%. (iii) Although the P value verification method based on field monitoring and simulation trial calculation has achieved satisfactory results in parameter selection, there are still some limitations in specific operations. For example, the collection of on-site monitoring data is easily affected by external factors. e modeling process needs to rely on assumptions. erefore, accurate collection of monitoring data and analysis of model influencing factors are the focus of future research.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request. Disclosure e preliminary results of this article have been published as a report manuscript of the "Southern African Transport Conference, Jul 2018." e link to the manuscript is as follows: http://www.satc.org.za/assets/2a_gao2.pdf.

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