Int J Civ Eng DOI 10.1007/s40999-017-0219-2

RESEARCH PAPER

Thermal Regime Analysis and Protective Measure Evaluation for Wide Embankment in Permafrost Regions of Qinghai-Tibet Plateau Tao Ma1 · Tao Tang1 · Xunhao Ding1 · Xiaoming Huang1 · Yongli Zhao1 

Received: 21 August 2016 / Revised: 19 February 2017 / Accepted: 27 March 2017 © Iran University of Science and Technology 2017

Abstract  This study analyzed the effects of a wide embankment on the thermal regime of permafrost and evaluated the feasibility of different measures to protect the permafrost underlying the wide embankment. Based on a specific embankment section of the Qinghai-Tibet highway and field observed data, a finite-element model was built and verified to conduct thermal analysis. Based on numerical analysis, the temperature fields and thawing depths of embankments with different widths were analyzed. The influences of different protective measures on the thermal stability of embankments were evaluated. The findings indicate that, after embankment construction, the annual ground temperature and maximum thawing depth increase with time. Thus, the embankment construction causes serious disturbance of the thermal stability of permafrost. An embankment with a larger width has a lower thermal stability. The results also show that raising the embankment height, setting a crushed stone layer, or setting an expanded polystyrene (EPS) layer can improve the thermal stability of an embankment. However, the improving effect of a single protective measure is not obvious for wide * Tao Ma [email protected] Tao Tang [email protected] Xunhao Ding [email protected] Xiaoming Huang [email protected] Yongli Zhao [email protected] 1



School of Transportation, Southeast University, 2 Sipailou, Nanjing 210096, China

embankments with widths larger than 26 m. A combination of a crushed stone layer and an EPS layer provides the best protective effect on the thermal stability of a wide embankment. Thus, it is recommended as a protective measure for wide embankments of expressways in permafrost regions of the Qinghai-Tibet Plateau. Keywords  Permafrost region · Thermal regime · Wide embankment · Protective measure

1 Introduction Permafrost is soil or rock with a temperature below 0 °C. It is distinguished from other geotechnical engineering materials due to the existence of ice and its sensitivity to temperature [1, 2]. The harsh and sensitive environmental conditions in permafrost regions give rise to great challenges for infrastructure construction and maintenance [3, 4]. For transportation infrastructure, such as railways and highways, the construction of embankments on permafrost may induce substantial disturbance in the heat and mass transfer balance between the ground surface and atmosphere, which results in more heat absorption in the embankments. This causes the temperature of the underlying permafrost to increase as the permafrost thaws, resulting in serious damage to embankments in the Qinghai-Tibet Plateau due to thaw settlement [5, 6]. Because of this effect, railway and highway construction has encountered challenges in permafrost regions for decades. It has been proven that highway construction in permafrost regions severely changes local conditions, such as the micro relief, surface conditions and physical properties of shallow ground [7]. This causes inevitable disturbance to the energy exchange between the atmosphere and the

13

Vol.:(0123456789)



Int J Civ Eng

ground, leading to thermal regime degradation of permafrost underlying highway embankments [8]. When asphalt pavement is used, the dark surface of the asphalt pavement absorbs a large amount of solar radiation, increasing the temperature of the pavement and negatively affecting the thermal regime of the underlying permafrost [9]. Since the physical and mechanical properties of permafrost are closely related to its thermal regime, the thermal stability degradation of the underlying permafrost can obviously undermine the stability and durability of highway embankments, which can severely affect the normal operation and service life of a highway [10]. Studies have shown that more than 60% of the underlying permafrost along the Qinghai-Tibet highway has experienced a noticeable temperature increase, permafrost table decline, and subsequent talik formation [11–13]. Observations have shown that structural failures of the highway have occurred progressively, which severely lowers the serviceability of the highway. Nearly 85% of these failures were due to roadbed thaw settlement induced by underlying permafrost warming and thawing [14–18]. To maintain the service performance of highways in permafrost regions, a series of engineering methods have been studied and practiced to protect the underlying permafrost and maintain the thermal balance between the atmosphere and permafrost. In general, the current protective approaches can be categorized into two types: heat-insulating approaches and heat-cooling approaches. Typical heatinsulating approaches mainly include raising the height of the subgrade, placing shading boards over the subgrade slope, using heat-insulating materials such as expanded polystyrene (EPS) in the subgrade, etc. [19–22]. Typical heat-cooling approaches mainly include crushed rock layers, ventilation ducts, and thermosyphons, etc. [23–25]. Although many of the methods have been practiced in permafrost regions and proven to be feasible for postponing the thermal degradation of permafrost, due to the complex environmental conditions of permafrost regions, pavement distresses caused by thawing and settlement of permafrost are still commonly observed in the field. Therefore, understanding the thermal regime and maintaining the thermal stability of permafrost under highways are still primary focuses of highway construction in permafrost regions. According to the National Expressway Network Plan of China, the Qinghai-Tibet Expressway (QTE), which is the first expressway connecting Tibet to other provinces in China, is to be constructed on the Qinghai-Tibet Plateau. It should be noted that expressways usually have wide embankments with design widths in the range of 24.5–45  m, which are approximately 2.5 to 4.5 times the current embankment width of the Qinghai-Tibet highway.

13

However, most of the previous research relating to the thermal regime and stability of permafrost has been based on narrow embankments, usually with design widths smaller than 20  m. Thus, to provide a fundamental basis for expressway construction in permafrost regions of the Qinghai-Tibet Plateau, this study analyzed the thermal regime of permafrost under wide embankments and evaluated the feasibility of different protective approaches for wide embankments.

2 Finite‑Element Modeling 2.1 Governing Equations and Equivalent Thermal Parameters By assuming no heat consumption of moisture evaporation in frozen soils, a two-dimensional heat-fluid coupling equations was proposed based on energy and mass conservation theory [26, 27], and can be described as follows: [ ] [ ] 𝜕T 𝜕T 𝜕T 𝜕 𝜕 C(T) 𝛽x (T) + 𝛽y (T) = (1) 𝜕t 𝜕x 𝜕x 𝜕y 𝜕y

C(T) = c𝜌 + 𝜌w

𝜕wu 𝜕T

𝛽x (T) = kx + L𝜌w Dx

(2)

𝜕wu 𝜕T

(3)

𝜕wu (4) 𝜕T where kx and ky are the components of soil thermal conductivity in the x- and y-directions, respectively (W m−1 K−1); Dx and Dy are the components of water diffusivity in the x- and y-directions, respectively ­(m2  s−1); ρ and ρw are the densities of soil and water, respectively (kg  m−3); c is the specific heat capacity of soil (J kg−1 K−1); C(T) is the equivalent volumetric heat capacity of soil (J m−3 K−1); L is the volumetric latent heat of thawing (J m−3); wu is the volumetric unfrozen soil moisture (­m3  m−3); βx(T) and βy(T) are the components of the equivalent thermal conductivity of soil at x and y direction, respectively (J m−1 s−1 K−1); T is the temperature (°C); t is time (s); and x and y represent the horizontal direction and vertical direction in the cross section of the embankment, respectively. As frozen soil is a composite material, the mineral and organic components in the frozen soil also contribute to its thermal properties [28]. Depending on its physical state (frozen or thawed) and the thermal properties of components, the equivalent thermal properties of frozen soil can be obtained in the following way:

𝛽y (T) = ky + L𝜌w Dy

Int J Civ Eng

C(T) =

{( ) csu + w0 cw 𝜌d

T ⩾ −𝜃f

( ) 𝜌d csf + 𝜌d ci w0 + 𝜌d cw − ci ⋅ w0 𝜃fb (−T)−b + L ⋅ w0 𝜃fb ⋅ b ⋅ (−T)−(b+1) {

𝛽x (T) = 𝛽y (T) =

T ⩾ −𝜃f

𝜆u 𝜆f + LD ⋅

w0 𝜃fb

⋅ b(−T)

−(b+1)

T < −𝜃f

(6) where csu and csf are the specific heat capacities of soil in the thawed and frozen states, respectively (J kg−1 K−1); w0 is the initial volumetric unfrozen soil moisture (­m3  m−3); ci and cw are the specific heat capacities of ice and water, respectively (J  kg−1  K−1); ρd is the dry density of soil (kg m−3); θf is the absolute value of the freezing temperature of soil; b is a constant related to soil type; λu and λf are the thermal conductivities of the thawed and frozen soil, respectively (W  m−1  K−1); and D is the water diffusivity ­(m2 s−1). 2.2 Geometric Section and Parameters Figure 1 shows a typical monitoring section of the QinghaiTibet highway, where the embankment with a width of 8 m and height of 2.5  m (shown as layer I in the figure) was filled with gravel soil and the natural ground can be divided into three layers (shown as layers II, III, and IV in the figure) with different types of soil, consisting of pebbly clay, gravel loam, and strong erosion mudstone. In the figure, the expanded polystyrene (EPS) layer and Crushed rock layer were not in the original monitoring section and they were shown for the further numerical modeling and analysis. This section was used as the basic computational region for the following numerical modeling and analysis. During

(5)

T < −𝜃f

the following numerical modeling, different height and width of embankment were considered. The EPS layer with a thickness of 0.1  m was set at a depth of 1  m below the pavement surface, and the crushed rock layer with a thickness of 1.5 m was set at a depth of 0.5 m below the natural ground surface. The specific material parameters for different layers shown in Fig. 1 are listed in Table 1. Table 2 summarizes the pavement structure and corresponding Table 1  Material parameters for different parts of the computational region Part

I

II

III

IV

Material

Gravel soil Pebbly clay Gravel loam Strong erosion mudstone ρd (kg/m3) 1800 1700 1300 1500 25 30 30 30 w0 (%) 0.84 0.84 0.84 csu (kJ/kg °C) 0.79 0.73 0.75 0.75 csf (kJ/kg °C) 0.71 4.182 4.182 4.182 cw (kJ/kg °C) 4.182 2.09 2.09 2.09 ci (kJ/kg °C) 2.09 1.95 0.87 1.47 λu (W/m °C) 1.919 2.69 1.22 1.82 λf (W/m °C) 1.980 3.73 3.44 D (J/m s) 9.35 × 10−6 4.66 −θf (°C) −0.20 −0.10 −0.19 −0.05 L (kJ/kg) 334.56 334.56 334.56 334.56 b 0.610 0.733 0.574 0.474

Fig. 1  Typical monitoring section of Qinghai-Tibet highway and setting illustration of EPS layer and crushed rock layer

13



Int J Civ Eng

Table 2  Pavement structure and corresponding material parameters

Layer

Upper asphalt layer

Lower asphalt layer

Base

Subbase

Cushion

Material

AC13

AC20

ATP25

Graded gravel

Thickness (m) Density (kg/m3) Specific capacity (J/kg °C) Thermal conductivity (J/m h °C)

0.04 2300 1670 4140

0.05 2320 1670 4320

0.12 2350 1260 2916

Cement stabilized macadam 0.24 2200 960 5616

Table 3  Thermal parameters of EPS and crushed rock 3

Materials

Density (kg/m )

Specific capac- Thermal ity (J/kg °C) conductivity (J/m h °C)

EPS Crush rock

40 1490

1400 839

108 1426

material parameters. The thermal parameters for the EPS and crushed rock are summarized in Table 3. 2.3 Numerical Modeling and Verification Based on the basic section shown in Fig.  1, a finite-element model was built based on ABAQUS. To analyze the influences of the width and height of the embankment on the thermal regime of the underlying permafrost, Table 4  Initial ground temperatures

Table 5  Monthly radiation data

Depth/m Temperature (°C) Depth (m) Temperature (°C)

Month

the width and height of the embankment in the numerical model were adjustable. An eight-nodded quadrilateral heat-conduction element was used in the numerical model. A UMATHT model was developed in ABAQUS to define equivalent thermal parameters. The initial ground temperatures used in the numerical model were taken from the field observed data collected in August. Table 4 shows the initial temperatures at different depths from the ground surface. In summer, the air temperature is higher than the soil temperature, and the soil absorbs heat from the outside environment, so the top soil has the highest temperature and the soil temperature decreases with increasing depth. Based on the field observed data, Table  5 summarizes the monthly average daily total solar radiation and effective ground radiation used in the numerical analysis. The albedo of asphalt pavement, side slope, and natural ground were 0.13, 0.3, and 0.3, respectively. The monthly average wind velocities are summarized in Table 6. Considering the effects of

0.5 6.11

1.0 2.62

1.5 0.55

2.0 −0.40

2.5 −1.35

3.0 −1.57

3.5 −1.86

4.0 −2.03

4.5 −2.03

5.0 −2.09

5.5 −2.12

6.0 −2.10

6.5 −2.15

7.0 −2.00

7.5 −2.25

8.0 −2.25

8.5 −2.25

9.0 −2.25

9.5 −2.25

10.0 −2.00

8

9

10

11

12

21.6

18.3

15.2

13

10.2

1

2

3

4

5

6

7

10.1 12.3 17.5 20.4 22.0 22.5 24 Daily solar radiation (MJ/m2 ­d−1) Effective 6.82 6.80 8.63 8.68 8.38 7.44 7.51 ground radiation (MJ/m2 d−1) Table 6  Monthly average wind velocities

13

0.2 2000 1100 6048

7.26

7.54

9.46 8.96

8.21

Month

1

2

3

4

5

6

7

8

9

10

11

12

Wind velocities (m/s)

5.69

5.83

5.92

4.79

4.37

4.11

3.72

3.37

3.38

3.58

4.29

5.23

Int J Civ Eng

global warming, the annual average air temperature will rise approximately 2.6 °C in the coming 50  years [29], and the monthly average temperature during numerical analysis can be described by Eq. (7). Finally, a userdefined DFLUX model and Film model were developed in ABAQUS to define the thermal boundary conditions for the numerical model. ) ( 2𝜋 5𝜋 2.6t T = −5.2 + 11.3 sin t+ + (7) 12 9 600 where T is the monthly average air temperature (°C); and t is time (month). Based on the specific embankment section shown in Fig.  1, numerical analysis was conducted to obtain its temperature field 10 years after the embankment was constructed. The field embankment was constructed in 2002. The temperature values of the field embankment were observed over a 10-year period. In addition, the monitoring points are set from depth 0 to 8 m at an interval of 0.5 m. Both the calculated temperatures and the field monitoring temperatures at different depths along the centerline of the embankment section are shown in Fig. 2. Since the numerical modeling parameters are simplified and fixed, while the field embankment parameters are complex and fluctuant, the temperature data from the numerical analysis are not exactly the same as the field observed data. However, they are similar to each other and have a similar variation trend. Therefore, the validity of the numerical modeling to investigate the thermal regime of the embankments in the permafrost regions was demonstrated.

3 Results and Discussion 3.1 Thermal Regime of Normal Embankment Section To analyze the influences of the embankment width on the thermal regime of the embankment section, numerical analysis was conducted on normal embankments with the same height of 4 m but with different widths, including 10, 18, 26, 34, 42, and 50 m. To analyze the influences of the embankment height on the thermal regime of the embankment section, a numerical analysis was conducted on normal embankments with different heights, including 2, 3, 4, 5, and 6 m. Figures  3 and 4 show the 10th-year temperature fields within the six different embankment sections in January and August, respectively. In the figures, when the temperature increases gradually, the colors turn from dark blue to dark red. Figure 3 shows that the embankment with a width of 10 m has no thawing core (represented by the dark red zone, with a temperature higher than 0 °C), while thawing cores were found for the embankments with widths of 18, 26, 34, 42, and 50 m. It can also be seen that the size of the thawing core apparently increases with an increase in the embankment width. Figure 4 shows that all of the embankments, along with part of the underlying ground, are in a thawing state during the hot season. It is also indicated from Fig.  4 that with an increase in embankment width, the 0 °C contour (shown as the edge of the dark blue zone, with temperature lower than 0 °C) moves down gradually, changes from a convex shape to a concave shape. This indicates that the thawing depth of natural ground underlying

Fig. 2  Comparison between calculated and measured temperature data

13



Int J Civ Eng

Fig. 3  Temperature fields in January for embankments with different widths: a 10 m, b 18 m, c 26 m, d 34 m, e 42 m, and f 50 m

the embankment increases with an increase in the embankment width. Figures  3 and 4 prove that greater embankment widths cause more serious disturbances to the temperature field of the embankment section. This is mainly because, compared with a narrow embankment, the intensity of heat absorption in a wide embankment will be higher because of the increased width of the black asphalt pavement. Meanwhile, it is more difficult for the heat at the center of the embankment to dissipate outwards due to the extended heat diffusion path. Thus, a wider embankment leads to faster permafrost degradation, as shown by the thawing depth increasing with an increase in the embankment width. The temperature variations during two decades at a depth of 5  m along the centerline of embankments with

13

different widths are shown in Fig.  5. It can be seen that the temperature variations with service time for all of the six embankments present similar trends, which shows that the temperature changes periodically over time and the average annual temperature continues to increase. It also indicates that, for all time points, a wider embankment has a higher temperature. Based on the temperature field analysis, the maximum thawing depth 10 years after embankment construction at the centerline and shoulder of embankments with different widths are summarized in Fig.  6. The thawing depths at the centerline of embankments with widths of 18, 26, 34, 42, and 50  m are 0.47, 2.29, 2.72, 3.12, and 3.77  m greater than that of an embankment with a width of 10  m, respectively. The thawing depth at the shoulder of embankments with

Int J Civ Eng

Fig. 4  Temperature fields in August for embankments with different widths: a 10 m, b 18 m, c 26 m, d 34 m, e 42 m, and f 50 m

widths of 18, 26, 34, 42, and 50  m is 0.52, 0.96, 1.15, 1.47, and 1.51  m greater than that of an embankment with a width of 10 m, respectively. It is clearly indicated that the thawing depth increases with an increase in the embankment width. Meanwhile, once the embankment width exceeds 18  m, the embankment has more influence on the thawing depth of the embankment centerline than on the thawing depth of the embankment shoulder. Therefore, a larger embankment width results in larger thawing depth and more seriously uneven thawing problems for the embankment. For embankments with different widths, the maximum thawing depths at the centerline and shoulder of the embankment versus the embankment height are shown in Fig. 7. In the figures, using the natural ground surface as

the zero line, a positive depth value represents that the maximum thawing depth is in the natural ground, while a negative depth value means that the maximum thawing depth is within the embankment. It can be seen that, for all of the embankments, the maximum thawing depth decreases with an increase in the embankment height. When the embankment height reaches 6  m, the maximum thawing depths at the centerline and shoulder of the embankments with widths smaller than 26  m are within the embankment or close to the natural ground surface, while the thawing depths of embankments with widths bigger than 18 m are still within the natural ground. The thawing depth difference between the centerline and shoulder versus embankment width is summarized in Fig.  8. It can be seen that the thawing depth difference clearly increases with an increase in the embankment

13



Int J Civ Eng

Fig. 5  Temperatures at depth of 5 m along the centerline of embankments with different widths

Fig. 6  Maximum thawing depth of embankments with different widths

width. The thawing depth differences between the centerline and shoulder of embankments with widths larger than 18 m are more obvious than those of embankments with widths smaller than 26 m. Therefore, it is indicated that increasing the embankment height is helpful for improving the thermal regime of the embankment section; however, the improvement is obviously weakened by increasing the embankment width, especially when the embankment width exceeds 18 m.

13

3.2 Influences of Different Protective Measures Three different protective measures, including a crushed stone layer, an EPS layer, and a combination of a crushed stone layer and an EPS layer, were used to improve the thermal regime of a wide embankment. Numerical analysis was conducted to compare the thermal regime of different embankments, including a normal embankment without any protective measures, an embankment with a crushed stone layer referred to as the CS embankment,

Int J Civ Eng Fig. 7  Influences of embankment height on the thawing depth: a at centerline and b at shoulder

Fig. 8  Thawing depth difference between centerline and shoulder

an embankment with an EPS layer referred to as the EPS embankment, and an embankment with a combination of a crushed stone layer and an EPS layer referred to as the composite embankment. Taking the embankment with a width of 50 m and height of 3  m for example, Figs.  9 and 10 show the tenth-year

temperature fields within the four different embankments in January and August, respectively. From Fig. 9, it can be seen that, even in the coldest season, an obvious thawed core (represented by the dark red zone, with temperature higher than 0 °C) is found within the normal embankment section and CS embankment section. However, the thawed

13



Int J Civ Eng

Fig. 9  Temperature fields in January for different embankments: a normal embankment, b CS embankment, c EPS embankment, and d Composite embankment

Fig. 10  Temperature fields in August for different embankments: a normal embankment, b CS embankment, c EPS embankment, and d composite embankment

core within the CS embankment section is much smaller than that within the normal embankment section. Meanwhile, there are no thawed cores within the EPS embankment and composite embankment. From Fig. 10, it can be seen that the 0 °C contours (shown as the edge of the dark blue zone, with temperature lower than 0 °C) of the normal embankment and CS embankment are shown within the natural ground, while the 0 °C contours of the EPS

13

embankment and composite embankment are shown within the embankment. Meanwhile, the 0 °C contour of the CS embankment is closer to the natural ground surface than that of the normal embankment, while the 0 °C contour of the EPS embankment is closer to the natural ground surface than that of the composite embankment. This proves that all of the protective measures can improve the thermal

Int J Civ Eng Fig. 11  Temperatures at depth of 5 m along the centerline of different embankments

regime of an embankment section; however, different protective measures have different improving effects. The temperature variations during two decades at a depth of 5  m along the centerline of different embankments are shown in Fig. 11. From Fig. 11, it can be seen that the temperature variations with service time for all of the four embankments present similar trends, showing that the temperature changes periodically over time and the average annual temperature continues to increase. For all time points, the normal embankment has the highest temperature, while the composite embankment has the lowest temperature. The EPS embankment has similar temperatures to the CS embankment before the 15th year after embankment construction and lower temperatures than the CS embankment after the 15th year after embankment construction. Based on the temperature field analysis, variations of the thawing depths at the embankment centerline along with service time for embankments with different widths are summarized in Fig.  12. The thawing depths 20  years after embankment construction at the centerline and shoulder of the embankment versus embankment width are shown in Fig. 13 for the different embankments. It is clearly indicated that, with the same embankment width and service time, the normal embankment has the biggest thawing depth, while the composite embankment has the smallest thawing depth. The thawing depth of the EPS embankment is smaller than that of CS embankment. With the embankment width increasing, the thawing depth of the normal embankment continues increasing rapidly, while the thawing depth of the composite embankment slowly increases. For the EPS and CS embankments, the thawing depth increases slowly with an increase in the embankment width until the embankment

width reaches 26  m and then increases rapidly with an increase in the embankment width once the embankment width exceeds 26  m. Meanwhile, as shown in Fig.  14, the thawing depth differences between the centerline and shoulder of the EPS and CS embankments become more obvious once the embankment widths exceeds 26  m. Therefore, the protective effects of the EPS layer and crushed stone layer can be obviously weakened by a wide embankment with a width exceeding 26  m, while the combination of an EPS layer and crushed stone layer has the best protective effects for the thermal regime of the wide embankment section.

4 Conclusions In this study, finite-element modeling was used and verified based on field data to investigate the thermal regime of wide embankments and evaluate different protective measures for wide embankments. The following conclusions can be drawn: 1. It is indicated that embankment construction in permafrost regions has serious influences on the thermal regime of the underlying permafrost ground. The temperature field of the embankment section shows annual periodic variation, and the annual average temperature continues to increase with an increase in the service time. This is the main reason for the degradation of permafrost underlying the embankment. 2. The results show that, for the same service time, an embankment with a larger width has a higher annual ground temperature and a greater maximum thawing depth. Meanwhile, a larger embankment width results

13



Int J Civ Eng

Fig. 12  Thawing depth at the embankment centerline vs service time for embankments with different widths: a 10 m, b 18 m, c 26 m, d 34 m, e 42 m, and f 50 m

in a more seriously uneven temperature field within the embankment section. Therefore, the thermal stability of an embankment section deteriorates with an increase in the embankment width. An expressway with wide embankments faces more serious thawing and uneven settlement problems than the current highway. 3. It is found that increasing the embankment height is helpful for improving the thermal regime of the embankment section. However, the improvement decreases with an increase in the embankment width.

13

When the embankment width exceeds 18  m, the improvement on the thermal regime of natural ground underlying the embankment is limited. Meanwhile, the determination of embankment height must consider many complex factors, such as mechanical properties of the ground and embankment, engineering effort, and financial cost. Therefore, other effective measurements must be considered to protect the thermal stability of the ground permafrost underlying expressways with wide embankments.

Int J Civ Eng Fig. 13  20th-year thawing depth vs embankment width for different embankments: a at centerline and b at shoulder

the National Science and Technology Support Program (2014BAG05B04), the Fundamental Research Funds for the Central Universities (No. 2242015R30027), and the Natural Science Foundation of Jiangsu Province (BK20161421).

References

Fig. 14  Thawing depth between centerline and shoulder vs embankment width for different embankments

4. Both an EPS layer and a crushed stone layer can improve the thermal regime of an embankment section. However, after the width of an embankment exceeds 26 m, the protective effects of an EPS layer or crushed stone layer are obviously weakened with the increase in width embankment. Since the objective of an EPS layer is to provide heat resistance for permafrost ground in hot seasons and a crushed stone layer aims to facilitate the heat dissipation of permafrost ground during cold seasons, the combination of an EPS layer and a crushed stone layer has the best protective effect for a wide embankment. 5. Further study will be conducted to investigate the thermal settlement of wide embankments for expressways and to evaluate other protective measures for expressway embankments. Acknowledgements  The study is financially supported by the National Natural Science Foundation of China (No. 51378006),

1. Ma T, Zhong Y, Tang T, Huang X (2016) Design and evaluation of heat-resistant asphalt mixture for permafrost regions. Int J Civ Eng 14:339–346. doi:10.1007/s40999-016-0039-9 2. Tang T, Zha X, Xiao Q, Chen Y (2016) Laboratory characterization and field validation of ROADMESH-reinforced asphalt pavement in China. Int J Civ Eng. doi:10.1007/ s40999-016-0128-9 3. Ma W, Cheng GD, Wu QB (2009) Construction on permafrost foundations: lessons learned from the Qinghai–Tibet railroad. Cold Reg Sci Technol 59(1):3–11 4. Zhang JW, Li J, Quan XJ (2013) Thermal stability analysis under embankment with asphalt pavement and cement pavement in permafrost regions. Sci World J 2013:1–12 5. Ma W, Mu Y, Wu Q, Sun Z, Liu Y (2011) Characteristics and mechanisms of embankment deformation along the Qinghai– Tibet railway in permafrost regions. Cold Reg Sci Technol 67(3):178–186 6. Tian Y, Fang J, Shen Y (2013) Research on EPS application to very wide highway embankments in permafrost regions. Sci Cold Arid Reg 5(4):503–508 7. Peng H, Ma W, Mu YH, Jin L (2015) Impact of permafrost degradation on embankment deformation of QinghaiTibet highway in permafrost regions. J Cent South Univ 22:1079–1086 8. Wu Q, Niu F (2013) Permafrost changes and engineering stability in Qinghai-Xizang Plateau. Chin Sci Bull 58(10):1079–1094 9. Li JP, Sheng Y (2008) Analysis of the thermal stability of an embankment under different pavement types in high temperature permafrost regions. Cold Reg Sci Technol 54:120–123 10. Zheng B, Zhang J, Qin Y (2010) Investigation for the deformation of embankment underlain by warm and ice-rich permafrost. Cold Reg Sci Technol 60(2):161–168 11. Sheng Y, Zhang JM, Liu YZ, Wu JM (2002) Thermal regime in the embankment of Qinghai–Tibetan highway in permafrost regions. Cold Reg Sci Technol 35(1):35–44

13



Int J Civ Eng

12. Wu QB, Liu YZ, Zhang JM, Tong CJ (2002) A review of recent frozen soil engineering in permafrost regions along Qinghai–Tibet highway, China. Permafrost Periglac Process 13(3):199–205 13. Wang S, Niu F, Zhao L, Li S (2003) The thermal stability of roadbed in permafrost regions along Qinghai–Tibet highway. Cold Reg Sci Technol 37(1):25–34 14. Liu YZ, Wu QB, Zhang JM, Sheng Y (2002) Deformation of highway roadbed in permafrost regions of the Tibetan Plateau. J Glaciol Geocryol 24(1):10–15 15. Wu QB, Mi HZ (2000) Predictions and control proposes of frozen ground under asphalt pavement at the high ground temperature section of the Qinghai–Xizang highway. Hydrogeol Eng Geol 27(2):14–17 16. Yu QH, Liu YZ, Tong CJ (2002) Analysis of the subgrade deformation of the Qinghai-Tibetan highway. J Glaciol Geocryol 24(5):623–627 17. Zhang C, Wang H, You Z, Ma B (2015) Sensitivity analysis of longitudinal cracking on asphalt pavement using MEPDG in permafrost region. J Traffic Transp Eng (English Edition) 2(1):40–47 18. Niu F, Cheng G, Ni W, Jin D (2005) Engineering-related slope failure in permafrost regions of the Qinghai-Tibet Plateau. Cold Reg Sci Technol 42(3):215–225 19. Wu Q, Zhang Z, Liu Y (2010) Long-term thermal effect of asphalt pavement on permafrost under an embankment. Cold Reg Sci Technol 60(3):221–229 20. Chen J, Sheng Y, Cheng GD (2006) Discussion on protection measures of permafrost under the action of engineering from the

13

point of earth surface energy balance equation in Qinghai–Tibet Plateau. J Glaciol Geocryol 28:223–228 21. Feng WJ, Ma W, Zhang LX, Wu ZJ (2003) Application of awning to roadway engineering in permafrost regions. J Geotech Geoenviron Eng 25:567–570 22. Wang SJ (2008) Highway construction technology in the permafrost region of China. China Communications Press, Beijing 23. Dong Y, Lai Y, Xu X, Zhang S (2010) Using perforated ventilation ducts to enhance the cooling effect of crushed-rock interlayer on embankments in permafrost regions. Cold Reg Sci Technol 62(1):76–82 24. Wu Q, Lu Z, Tingjun Z, Ma W, Liu Y (2008) Analysis of cooling effect of crushed rock-based embankment of the Qinghai–Xizang railway. Cold Reg Sci Technol 53:271–282 25. Pan Y, Wu CZ (2002) Numerical investigations and engineering applications on freezing expansion of soil restrained two-phase closed thermosyphons. Int J Therm Sci 41:341–347 26. Harlan RL (1973) Analysis of coupled heat-fluid transport in partially frozen soil. Water Resour Res 9(5):1314–1323 27. Taylor GS, Luthin JN (1978) A model for coupled heat and moisture transfer during soil freezing. Can Geotech J 15(4):548–555 28. Farouki OT (1981) The thermal properties of soils in cold regions. Cold Reg Sci Technol 5(1):67–75 29. Liu J, Tian Y (2002) Numerical studies for the thermal regime of a roadbed with insulation on permafrost. Cold Reg Sci Technol 35(1):1–13