Fiber optic monitoring of an anti-slide pile in a retrogressive landslide

Lei Zhang ,Honghu Zhu ,Heming Han ,Bin Shi

a School of Water Resources and Environment,China University of Geosciences,Beijing,100084,China

b State Key Laboratory of Hydroscience and Engineering,Tsinghua University,Beijing,100084,China

c School of Earth Sciences and Engineering,Nanjing University,Nanjing,210023,China

d School of Civil and Environmental Engineering,Nanyang Technological University,639798,Singapore

Keywords: Anti-slide pile Multi-sliding surface Pile-soil interface Brillouin optical time domain reflectometry(BOTDR)Geotechnical monitoring Reservoir landslide

ABSTRACT Anti-slide piles are one of the most important reinforcement structures against landslides,and evaluating the working conditions is of great significance for landslide mitigation.The widely adopted analytical methods of pile internal forces include cantilever beam method and elastic foundation beam method.However,due to many assumptions involved in calculation,the analytical models cannot be fully applicable to complex site situations,e.g.landslides with multi-sliding surfaces and pile-soil interface separation as discussed herein.In view of this,the combination of distributed fiber optic sensing (DFOS) and strain-internal force conversion methods was proposed to evaluate the working conditions of an anti-sliding pile in a typical retrogressive landslide in the Three Gorges reservoir area,China.Brillouin optical time domain reflectometry (BOTDR) was utilized to monitor the strain distribution along the pile.Next,by analyzing the relative deformation between the pile and its adjacent inclinometer,the pile-soil interface separation was profiled.Finally,the internal forces of the anti-slide pile were derived based on the strain-internal force conversion method.According to the ratio of calculated internal forces to the design values,the working conditions of the anti-slide pile could be evaluated.The results demonstrated that the proposed method could reveal the deformation pattern of the anti-slide pile system,and can quantitatively evaluate its working conditions.

1.Introduction

Due to the construction of the Three Gorges Dam,a large number of landslides have been induced and revived as a result of the dynamic change of hydrological factors (Gu et al.,2017;Zhang et al.,2018;Tang et al.,2019;Yin et al.,2022).The reservoir landslides will directly threaten lives and properties,and trigger secondary disasters(e.g.surges,debris flows,and barrier lakes),which poses a severe threat to downstream roads,bridges,and other infrastructures (Crosta et al.,2016;Iba?ez and Hatzor,2018).Antislide piles,as an important retaining structure,are widely used for stabilizing slopes(Ito and Matsui,1975;Li et al.,2017;Sánchez and Roesset,2013).Performance evaluation of anti-slide piles is of critical importance for landslide risk mitigation.

It is well-known that internal forces of anti-slide piles are closely related to the working condition (Liu et al.,2020).Many internal force calculation methods have been proposed,among which the cantilever beam method and elastic foundation beam method are most widely used.For the cantilever beam method as illustrated in Fig.1a,the sliding surface of the landslide is generally assumed as the boundary of the cantilever beam and elastic foundation beam.Specifically,the anti-slide pile above a sliding surface is assumed to be a cantilever beam.Below the sliding surface,it is regarded as an elastic foundation beam(Zheng et al.,2010).Based on this,it can be found that the cantilever beam method cannot be fully applicable for landslides with multi-sliding surfaces.For the elastic foundation beam method(see Fig.1b),the whole pile is assumed to be an elastic foundation beam regardless of the sliding surface(Zhu et al.,2021).In this case,the anti-slide pile is subjected to the thrust force provided by the soil at the rear edge and the resistance provided by the soil at the front edge.Given this,the elastic foundation beam cannot be used for the case of pile-soil interface separation.This is because the continuous separation of the pile and soil will lead to a variation in the contact area between the soil and anti-slide pile(see Fig.1b),resulting in a change of resistance provided by the soil in front of the pile.Moreover,if there is an undetected sliding surface below the bottom of the anti-slide pile,it will lead to the overall sliding of the anti-slide pile.In this case,neither cantilever beam nor elastic foundation beam is suitable for calculating the internal force of antislide piles.Generally,the anti-slide piles interact with the landslide body and sliding bed,where the working conditions are affected by a variety of factors.When using monitoring data,its stability condition can be evaluated truly and accurately.

Fig.1.(a) Cantilever beam method;and (b) Elastic foundation beam method.

In recent years,the distributed fiber optic sensing (DFOS) technologies have been widely used in geohazards monitoring,which features anti-electromagnetic interference,distributed sensing,and multi-physical monitoring (Schenato et al.,2017;Minardo et al.,2018;Zhang et al.,2020a,2020b;Cao et al.,2022;Zhu et al.,2022a,2022b;Ye et al.,2022).These technologies have been applied to deformation monitoring of different types of pile foundations (Klar et al.,2006;Zhu et al.,2012;Wang et al.,2021).By adopting Brillouin optical time domain reflectometry (BOTDR)technology,continuous deformation measurement of the antisliding pile over elapsed time was realized by Xiao et al.(2019).The horizontal displacements of a multi-anchor pile wall during excavation were captured by a BOTDR-based system,which demonstrates that the fiber optic sensors can obtain deformation measurements with high accuracy (Moffat et al.,2019).In the study of Han et al.(2021),the deformation pattern of a deep diaphragm wall was revealed by analyzing the monitoring data recorded by an ultraweak fiber Bragg grating (FBG) interrogator.However,the above studies were carried out under simple geological conditions.For complex site conditions,e.g.landslides with multi-sliding surfaces or undergoing pile-soil separation,the deformation process and evolution mechanism of anti-slide piles have not been well investigated.In addition,previous studies are mainly focused on the foundation deformation characteristics,but the internal forces were rarely explored.More attentions should be given to this issue to get a better understanding of the working condition of anti-slide piles.

In this study,a monitoring system of DFOS-based anti-slide pile was established and long-term monitoring was conducted at the Majiagou landslide,a retrogressive landslide located in the Three Gorges reservoir area,China.By analyzing the relative movements of an anti-slide pile and its adjacent inclinometer,the pile-soil separation process is revealed.Based on the strain monitoring data,the internal force and deflection distribution along the pile are derived.Finally,the working condition of the anti-slide pile is analyzed and evaluated.

2.Background of the study area

The Majiagou landslide is situated in Guizhou Town,Zigui County,Hubei Province,China.It is located on the left bank of the Zhaxi River (Fig.2),a tributary of the Yangtze River.The distance between the landslide and the estuary of the Yangtze River is 2.1 km approximately(Ma et al.,2016).It is roughly tongue-shaped,whose main sliding direction is 290°(Fig.2).The length and width of this landslide are approximately 560 m and 150 m,respectively.The front edge is immersed by the reservoir water level all year,and the elevation is 110 m above sea level(Fig.3).The elevation of the rear edge is 280 m.The longitudinal profile ranging from 10°to 20°,with a relatively gentle slope,is characterized by horizontal terraces and steep scarps.The vertical profile of the landslide is illustrated in Fig.3.It is mainly composed of superficial deposits and sedimentary bedrock.The sedimentary bedrock mainly consists of interbedded sandstone,thin purple-red mudstone,and superficial deposits (Zhang et al.,2020b;Liao et al.,2021).

Fig.2.Location and topographic map of the Majiagou landslide.

Fig.3.Geological profile of the Majiagou landslide.

Since the impoundment of the Three Gorges Reservoir in 2003,this reservoir landslide began to deform.To combat the induced deformation,two anti-slide piles with lengths of 40 m were installed at an elevation of 200 m in 2012 to reinforce the landslide,as illustrated in Fig.3.Note that the Majiagou landslide is a typical retrogressive landslide (Zhang et al.,2018;He et al.,2020;Wang et al.,2023),whose deformation accumulates at the toe first and then involves the middle and top areas.Therefore,separation of the pile-soil interface may occur in the front edge of the pile.Given this,inclinometer B3 for monitoring the landslide deformation was installed in the vicinity of the anti-slide piles in 2012.By plotting the relative deformation of the inclinometer and anti-slide pile,the evolution process of the pile-soil separation could be profiled.

3.Methodology

In this study,the BOTDR technology was utilized to monitor the strain distribution along the pile.By integrating the strain difference on both sides of the anti-slide pile,the pile deformation data could be obtained.Finally,to achieve the working condition evaluation,the internal force of the anti-slide pile was derived based on the strain-internal force conversion method.

3.1.Principle of brillouin optical time domain reflectometry(BOTDR)

The strain distribution along a fiber optic cable can be recorded using BOTDR technology.When a laser pulse travels in an optical fiber,the Brillouin scattered light will return for interpretation.The frequency of this scattered light will shift in terms of the temperature and axial strain along the fiber.A linear relationship between the Brillouin frequency shift (BFS),temperature,and strain is as follows (Ohno et al.,2001):

where vB（ε,T）is the BFS when the temperature isTand strain is ε;vB（0,T0）is the BFS without strain at temperatureT0;C1is the strain coefficient,which is approximately 0.05 MHz/με;andC2is the temperature coefficient,which is approximately 1 MHz/°C (Zhu et al.,2014;Zhang et al.,2021).According to Eq.(1),after temperature compensation,the strain along a fiber optic cable can be continuously measured.

3.2.Lateral deflection calculation based on BOTDR

The lateral deflection calculation method of the anti-slide pile based on BOTDR is depicted in Fig.4.The distributed strain sensing(DSS)cables are attached to the steel bar at the front and rear edge of the anti-slide pile (see Fig.4a).The anti-slide pile will undergo bending under the action of landslide thrust.According to standard engineering beam theory,the relationship between strain εm（z）induced by bending at any pointy（z）,curvature radius ρ（z）and deflection ω（z）can be described as follows(Mohamad et al.,2012):

Fig.4.Deformation monitoring principle of the anti-slide pile with BOTDR:(a)Layout of DSS cables;and (b) Diagram of lateral deflection calculation of the anti-slide pile.

whereDis the distance between two opposite DSS cables,Eis the elastic modulus,Iis the area moment of inertia,Mis the bending moment,dzis the length of the segment,dω（z）is the deflection variation with the length of dz,ρ（z）is the curvature radius of the pile,εu（z）is the strain at the front edge of the pile,and εd（z）is the strain at the rear edge of the pile.

Owing to the high shear strength of reinforced concrete,the angle is generally small when the anti-slide pile undergoes bending.Therefore,dω（z）/dzcan be ignored because it is far less than 1 (Hu et al.,2017).Therefore,Eq.(3) can be modified as follows:

Considering that the anti-slide pile is embedded in the deep bedrock of the Majiagou landslide,it suggests that the bottom can be regarded as fixed boundary.Based on such boundary conditions,the following equation can be obtained:

3.3.Internal force calculation based on BOTDR

According to the mechanics of materials,based on the strain data,the following formulas can be used to calculate the internal force of anti-slide piles (Zou et al.,2020).

where εm（z）is the strain recorded by the DSS cable at depthz,M（z）is the bending moment of the anti-slide pile,andF（z）is the shear force of the anti-slide pile.

By substituting the recorded strain data into Eqs.(6)and(7),the distribution of the bending moment and shear force can be obtained.

4.Monitoring system of DFOS-based anti-slide pile

4.1.Monitoring scheme of the anti-slide pile

The monitoring system of the DFOS-based anti-slide pile deformation is depicted in Fig.5.DSS cables and distributed temperature sensing (DTS) cables,with detailed parameters listed in Table 1,are used to record the strain and temperature variation along the pile.The pile length is 40 m,and the cross section of the pile is 1.5 m × 2.0 m,as shown in Fig.6.The DSS cables were fastened along the steel bar within the anti-slide piles.The DSS cables B1-B2 and D1-D2 in the “U” shape were fastened with binding wire at the opposite side of the anti-slide pile against the sliding direction.DTS cables A1 and C1 were installed at adjacent sides of the pile in a stress-free state(Figs.5 and 6).By connecting the N8511 BOTDR demodulator,whose sampling interval and measurement accuracy are 5 cm and±40 με,the strain distribution along the DSS cables at the opposite side of the anti-slide pile can be measured.Likewise,the temperature distribution along DTS cables can be captured by connecting the NZS-DTS-M06 DTS demodulator,whose measurement accuracy is 0.5°.The monitoring of the anti-slide pile started in November 2011.As of July 2017,15 sets of data had been recorded.

Table 1Parameters of DSS and DTS cables used in this study.

Fig.5.DFOS-based deformation monitoring system for the anti-slide pile.

Fig.6.Layout of fiber optic cables in the anti-slide pile.

4.2.Monitoring results and analyses

After temperature compensation,the strain profiles exerted on cables B2 and D2 are illustrated in Fig.7a and b,respectively.The strain distribution at the front and rear edges was basically symmetrical.In view of this,the strain of the rear edge was selected for analysis.The strain peaks appeared at depths of 14 m,22 m and 27 m,respectively.As large strain measurements indicate that these areas underwent significant deformation,they could be used to identify the positions of sliding surfaces(Zhu et al.,2014;Zhang et al.,2018).Based on the strain monitoring data,three sliding surfaces,namely S1,S2,and S3,were identified and outlined in Fig.7b.By comparing them with the geological profile of the Majiagou landslide (Fig.5),sliding surface S1 was found to be located at the contact face between the superficial deposits and sedimentary bedrock.Two deeper sliding surfaces,i.e.S2 and S3,corresponded to the weak layers (mudstone) interbedded in the bedrock.

Fig.7.Strain distribution along the anti-slide pile:(a)At the front edge;and(b)at the rear edge.

Different from traditional anti-slide piles,the piles of the Majiagou landslide demonstrated a complex deformation pattern due to the presence of multi-sliding surfaces.Based on the recorded strain variation,the deformation pattern of the anti-slide pile is presented in Fig.8.Q1,Q2andQ3denote the thrust forces provided by the sliding mass along sliding surfaces S1,S2,and S3,respectively,whileP1,P2,andP3represent the resistance forces in front of the pile along the corresponding sliding surfaces,andPdis the reaction force at the bottom of the embedded section of the pile.As shown in Fig.7b,at depths ranging from 19 m to 34 m,the strain at the rear edge of the pile was positive,indicating that the pile bent forward along sliding surfaces S2 and S3 under the thrust forces ofQ2andQ3.The negative strain distribution at depths from 8 m to 19 m suggested that the pile underwent backward bending along sliding surface S1(Fig.8).This occurred because the deformation of the Majiagou landslide was mainly governed by the deep sliding surfaces S2 and S3,while the deformation of shallow sliding surface S1 was relatively small.Therefore,the upper part of the anti-slide pile experienced backward bending.Overall,due to the presence of multi-sliding surfaces,the pile exhibited an“S-shaped”bending.

Fig.8.Deformation mode of the anti-slide pile.

According to Eq.(5),the deflection curves of the anti-slide pile can be derived.As shown in Fig.9a,three sliding surfaces (S1,S2 and S3),located at depths of 14 m,21 m,and 27 m,respectively,can be identified.The locations of the sliding surfaces are consistent with the aforementioned analyses.The time-history deformation at the pile top is plotted in Fig.9b.It should be noted that sharp increases in the pile deformation appeared in July 2015 and July 2016,respectively.Before March 2015,the pile deformation was at a low level,with an annual average displacement rate of about 21.5 mm/y.However,after March 2015,the displacement significantly increased with time,reaching an annual average displacement rate of 59.4 mm/y,which was approximately 3 times the previous one.The rapid increase of anti-slide pile deformation indicated an unstable state of the landslide and anti-slide pile system.

Fig.9.(a) Deflection of the anti-slide pile;and (b) time-history displacement at the pile top.

As illustrated in Fig.10b,the depth corresponding to maximum strains along the pile moved from 20.5 m to 22 m with time elapsed.A similar phenomenon was also noticed by Hussien et al.(2010),which was induced by the pile-soil separation due to lateral loading.To confirm this,the relative movement between the anti-slide pile and its adjacent inclinometer was analyzed.Inclinometer B3 was installed right ahead of the anti-slide pile,as shown in Fig.3,and thus any pile-soil separation could be detected by their relative deformation.The deformation of the inclinometer was reported by Zhang et al.(2020a) and the deformation of the pile was illustrated in Fig.9a.

Fig.10.(a) Strain variation along the pile at the rear edge;and (b) maximum strain variation characteristics at sliding surface S2.

The displacement curves of the inclinometer relative to the antislide pile are depicted in Fig.11.In December 2012,a comparison of the pile deformations and inclinometer readings indicated that there was negligible gap between the pile surface and the surround soil.With elapsed time,a gap began to appear and develop downward.In September 2013,significant relative movements between the inclinometer and anti-slide pile were captured and the largest gap was observed at a depth of 15 m.By July 2015,the gap extended to a depth of 21 m and penetrated into sliding surface S2(Fig.11).Subsequently,the gap continued to go downward and reached a depth of 25 m two years later.During this period,obvious cracks on the ground surface between the pile and soil were witnessed.In September 2013,a visible crack between the soil and pile appeared and the width of the gap was approximately 3 cm (see Fig.12a).Subsequently,as shown in Fig.12b and c,the crack width expanded to 40 mm in July 2015.It should be noted that until July 2017,the crack width was still 40 mm.This indicated that the deformation of the soil in front of the pile was synchronous with that of the anti-slide pile.According to the study of Liao et al.(2021),the shear strength of gravel soil,which is the main components of sliding body in Three Gorges Region,deteriorates significantly due to soaking.Specially,the reduction in strength parameters,i.e.the cohesion and friction angle,can reach as high as 60% after 7 wetting-drying cycles.Therefore,it is inferred that the continuous sliding of soil mass in front of the pile can be attributed to the weakening of the soil strength due to periodic fluctuation of the reservoir water level.

Fig.11.Development of the gap between the soil and pile.

Fig.12.Development of the gap between the soil and pile.

The widths of surface cracks observed in the field are consistent with the monitoring results.Unfortunately,only the evolution of surface cracks was observed,while the process of pile-soil separation below the surface could not be profiled by field investigation.However,in general,the high precision of DFOS technology and traditional inclinometers in landslide monitoring had been demonstrated by a large number of studies (e.g.Simeoni and Mongiovì,2007;Stark and Choi,2008;Calcaterra et al.,2012;Zeni et al.,2015),which can achieve deformation measurements with millimeter level accuracy (Xiong and Zhang,1990;Sun et al.,2016;Zhang et al.,2020c,d).Therefore,the process of underground pile-soil separation outlined by analyzing the relative deformation of the soil and anti-slide pile is reasonably accurate.

5.Working condition evaluation of anti-slide piles

5.1.Impact of hydrological factors on anti-slide pile

The intense rainfall and 30 m fluctuation of the reservoir water level,which govern the seepage field change,are the key factors controling the deformation of reservoir landslides(Xia et al.,2013;Gao et al.,2018;Yao et al.,2019).To explore the influence of hydrological factors on the working condition of the anti-slide pile,the relationship between daily rainfall,cumulative rainfall,reservoir level fluctuation and deflection of anti-slide piles was established.

The deformation of the anti-slide pile in response to hydrological factors is illustrated in Fig.13.The pile deformation increased obviously from March to August each year.Especially from March to July 2015,the displacement of the pile marked with the yellow rectangle frame increased 55 mm.Considering that both rainfall and reservoir water level changes can induce the deformation of anti-slide piles by varying the seepage field inside the landslide,the relationship between rainfall,reservoir water level fluctuations and deformation was analyzed separately.In terms of rainfall,as shown in Fig.13d and e,the rainfall was concentrated as marked by the red circle;however,the deformation of the anti-slide pile did not show an increase subsequently.Specifically,intensive rainfall occurred in September 2013,September 2014 and July 2016,while the pile displacements remained almost unchanged.Considering that it took time for rainfall to infiltrate,the deformation of the anti-slide pile was usually hysteretic.However,even in a long period of time after rainfall,the pile remained non-deformed.It could therefore be concluded that rainfall was not the key factor that governed the deformation of the pile(Zhu et al.,2012).In terms of the fluctuation of the reservoir water level,the increase in anti-slide pile deformation corresponded to the decrease in the reservoir water level,which was consistent with previous studies (e.g.Jiang et al.,2011;Pinyol et al.,2012;Song et al.,2018).

Fig.13.Deformation of the anti-slide pile in response to the water level fluctuation and rainfall: (a) The displacement curve;(b) The peak strain recorded by DSS cables;(c) Fluctuation of reservoir water level;(d) Cumulative rainfall and (e) Daily rainfall.

For further analyses,the displacement rate relative to the average change in the reservoir water level is plotted in Fig.14.The positive or negative average change rate of the reservoir level represents that the water level rises or drops.It shows that the deformation of the anti-slide pile is negatively correlated with the fluctuation of the reservoir water level.That is,when the reservoir water level drops,the anti-slide pile undergoes significant deformation.In contrast,when the reservoir water level rises,the antislide pile hardly deforms.

Fig.14.Relationship between the displacement rate and average change in the reservoir water level.

5.2.Safety evaluation of the anti-slide pile

As mentioned above,the working conditions of anti-slide piles are of great importance for the stability evaluation of the Majiagou landslide.The factor of safety of a landslide refers to the ratio of the sliding resistance force to the thrust force(Sarma,1973).When the ratio is greater than 1,the landslide is stable.When it equals 1,the landslide is in the limit equilibrium state;and when it is less than 1,instability of landslide is induced.Inspired by this,Kis introduced and defined as the factor of safety of the anti-slide pile,which equals the ratio of the maximum internal force to the design value:

whereMdandQdare the designed bending moment and the shear force of anti-slide pile,respectively;andMmaxandQmaxare the maximum bending moment and the shear force measured by the DSS cables,respectively.

WhenK＞1,the designed internal force is greater than the measured value,signifying that the anti-slide pile works well;whenK=1,the anti-slide pile is in a critical state,which requires close monitoring;whenK＜1,failure or local failure of the antislide pile may occur,and reinforcement measures should be carried out.The internal force is calculated based on the measured strain data herein.The bending moment and shear force along the antislide pile can be derived according to Eqs.(6) and (7) and the results are outlined in Fig.15.The maximum bending moment and the shear force of the pile are 15011.8 kN m and 5192.7 kN and registered at 27 m and 25.6 m,respectively.

Fig.15.Distribution of bending moment and shear force along the anti-slide pile: (a) Bending moment;(b) Shear force;(c) Variation of bending moment with time at different depths;and (d) Variation of shear force with time at different depths.

By assuming that the Majiagou landslide is in the limit equilibrium state,the design value of anti-slide pile thrust for the landslide can be derived,which is 1062.66 kN/m(Yong,2014).The pile spacing is 6 m.The cross section of is 1.5 m (a) × 2.0 m (b)(length × width).Thus,the area of the pile cross section isab=3 m2,and the calculated width isBP=2.5 m.The pile moment of inertia isI=ba3/12=1 m4,and the elastic modulus of concrete isE=3×107kPa.Thus,the bending moment and the shear force of the anti-slide pile can be calculated to be 9314.1 kN and 86,924.4 kN m,respectively.Based on Eq.(8),the factor of safety of the anti-slide pile can be calculated as follows:

The results indicate that the bending moment of the anti-slide pile still has a certain safety margin,while the shear force is close to the design value.It is suggested to install prestressed anchors at the rear edge of ant-slide piles or embed mini anti-slide piles to bear part of the landslide thrust force(Zheng et al.,2010),reducing the shear force acting on the anti-sliding pile.

6.Discussions

The advantages of the proposed method in this study are summarized and some considerations for field applications are discussed in this section.

6.1.Advantages of the proposed method

Based on the strain/displacement measurements,the internal force of anti-slide piles can be derived (Ren et al.,2008;Hu et al.,2017).Strain gauges and inclinometer,as the most common strain/deformation monitoring instruments,are widely used in monitoring anti-slide piles.However,it is known that the maximum internal force is an important index to evaluate the stability condition of antislide piles.Strain gauges/inclinometer only provide limited number of data (Zhu et al.,2014),which is somewhat difficult to record the detail deformation information of the anti-slide pile.This will lead to misjudgment of working condition of anti-slide pile.As a contrast,the DFOS technology,with the unique advantages of long distance and distributed monitoring,can obtain the continuous strain distribution along anti-slide piles.Based on the strain-internal force conversion method,the bending moment and the shear force of antislide piles can be accurately derived.Moreover,the factor of safety of anti-slide pile is proposed.Through comparing the factor of safetyKto unit value,the safety margin of the bending moment and the shear force can be derived.Thus,the working condition of anti-slide pile can be evaluated quantitatively.In addition,according to the factor of safety,the risk grade of anti-slide pile can be divided.Based on different risk levels,corresponding treatment measures can be taken.

6.2.Field application

6.2.1.Selection of fiber optic cables

Due to the harsh geological environments,bare optical fibers are easy to break,which makes them unsuitable for field monitoring.Given this,a specially designed DSS cable is developed.As illustrated in Fig.16,the DSS cable,from inside to outside,consists of bare silica fiber optic,PVC sheath,steel wire mesh,and polyethylene (PE) jacket.Among them,the steel wire mesh can significantly improve the strength of the DSS cables,which greatly improves its survival rate in field application.

Fig.16.Specially designed fiber optic cables: (a) DDS cable and (b) DTS cable.

Since DFOS monitoring results are easily affected by temperature,temperature compensation should be conducted.In this study,the DTS cable,fabricated by Suzhou NanZee Sensing Technology Co.Ltd.,is selected as temperature sensing cable and installed along the anti-slide pile.As depicted in Fig.16b,the temperature sensing cable is composed of five layers.They,from inside to outside,are bare silica fiber optic,steel conduit,Kevlar fiber,steel wire mesh and polyethylene(PE) jacket.

6.2.2.Installation of fiber optic cables

For an anti-slide pile,the main rebar could bear most of thrust forces induced by landslide.Therefore,the DSS cables were attached to the main rebars so as to record the stress and deformation information of anti-slide pile.It needed to point out that when installing the DSS cables,a heavy weight should be placed at the bottom of the cable to ensure that the cable is always in a straight state during the whole installation process.Meanwhile,the DSS cables and the main rebars were bound together by binding steel wire,as illustrated in Fig.17.After grouting,the DSS cables“solidified” in and deform synchronously with the anti-slide pile.As temperature sensing cables were only sensitive to the temperature change,they were installed at adjacent sides of the pile in a stress-free state.

Fig.17.Field installation of DSS cables.

6.3.Treatment measures for landslides with pile-soil separation

As illustrated in Fig.18,the pile-soil separation phenomenon occurs when the soil mass in front of the anti-slide pile slides continuously towards the reservoir,resulting in development of the gaps between the pile and soil.During the rainy season,rainfall infiltrates into the gaps,forming new dominant seepage channels that contribute to the activation of the landslide.In view of this,the gaps between the anti-slide pile and soil need to be backfilled and it is highly recommended to use the impervious clay as the backfilling materials to prevent rainfall infiltration.

Fig.18.Treatment measures for landslides with pile-soil separation.

As mentioned above,the mechanical properties of the soil mass are weakened due to the long-term fluctuation of reservoir water level,which results in the continuous sliding away of the soil mass in front of the anti-slide pile.Given this,it is an effective way to install prestressed anchors in the front of landslide to prevent its continuous deformation (Fig.18).

7.Conclusions

In this study,the DFOS technology was utilized to record the strain distribution along the anti-slide pile in a typical retrogressive landslide.Some findings related to the working condition of antislide piles have been addressed as follows:

(1) DFOS technology enables high accuracy measurement of strain profiles along the anti-slide pile.By analyzing the distributed strain measurements,three sliding surfaces of the landslide have been identified,and the deformation mode and associated evolution process of the anti-slide pile have been revealed.

(2) The strain variations along the anti-slide pile indicate a gradual separation of the pile-soil interface over time,as verified by the relative movements of the anti-slide pile and its adjacent inclinometer.

(3) The field monitoring results reveal that the deformation of the anti-slide pile is negatively correlated with the fluctuation of the reservoir water level,whereas rainfall has minor influence.

(4) By analyzing the recorded strain data,the bending moment and shear force of the anti-slide pile can be derived.By comparing with the design internal forces,the results indicate that the bending moment still has a certain safety margin,while the shear force is approaching the design value.To migrate potential georisks,the use of mini antislide piles or installing prestressed anchors at the rear edge of the ant-slide pile is recommended.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors gratefully acknowledge the financial support provided by the Young Scientists Fund of the National Natural Science Foundation of China (Grant No.41907232),the National Science Fund for Distinguished Young Scholars of China (Grant No.42225702),and the State Key Program of National Natural Science Foundation of China (Grant No.41230636).