Experimental Study on Interface Frictional Characteristics between Sand and Steel Pipe Jacking

: In order to study the variation law of shear frictional characteristics of the steel pipe jacking and sand interface under different working conditions, the shear stress–strain curve between ﬁve different particle sizes of sand and steel pipe jacking under different normal stress and slurry lubrication conditions was measured by using a direct shear device, and the internal friction angle, friction coefﬁcient and cohesion of the pipe–soil interface were calculated by regression analysis. The test results show that the shear stress between sand and steel pipe jacking decreases with the increase of the average particle size of the sand, and the strain-softening phenomenon occurs. The normal stress does not change the trend of the shear stress–strain curve at the pipe–soil interface, and the peak and residual values of the shear stress increase with the increase of the normal stress. The peak and residual values of the shear stress at the pipe–soil interface under the slurry lubrication condition are smaller than those under the no slurry lubrication condition. The peak shear stress between the pipe and soil under the lubricated slurry condition decreases by about 20%. The internal friction angle and friction coefﬁcient of the pipe–soil interface decrease with the increase of the particle size, and there is no obvious pattern between the cohesion quantity relationship and the average particle size.


Introduction
Steel pipe jacking construction has high precision, good safety and environmental friendliness, and it is commonly used for underground pipeline laying.The principle of steel pipe jacking is to jack the pipe jacking machine into the stratum through hydraulic jacks.During the tunnelling process, the pipe jacking machine cutterhead cuts the soil in front, and the soil cut down is discharged through the exhaust-earth device.The prefabricated steel pipe is jacked into the stratum section by section under the action of jacking force, so as to complete the pipeline laying [1].In the process of steel pipe jacking construction, the pipe section in the jacking and the surrounding soil friction generate side friction, and the side friction will lead to insufficient jacking force; at the same time, the side friction resistance will disturb the soil.Serious cases induce soil-body collapsing, causing surface settlement, and the stability of the pipeline and the surrounding buildings and structures produce harm [2].Therefore, reducing the side friction between the pipe and soil can effectively ensure the safety of construction at present, often through the injection of bentonite slurry between the pipe section and the surrounding strata in order to change the contact mode between the pipe and soil, reducing the sliding friction coefficient between the pipe and soil so as to reduce the side friction [3].
The study of the friction of the steel pipe jacking wall is often referred to as concrete pipe jacking, which is often studied by scholars at home and abroad through indoor model tests, direct shear tests and field data analysis.Shou et al. [4] studied the frictional characteristics of bentonite slurry commonly used in pipe jacking construction in Taiwan by designing model tests, and the test results showed that bentonite slurry can reduce the friction coefficient of the interface of pipe and soil by 23%, and bentonite slurry with Appl.Sci.2023, 13, 2016 2 of 10 plasticizer can reduce the interfacial friction coefficient by 63%.Namli and Guler [5] studied the influence of slurry pressure size on the lubrication effect by changing the injection pressure.The test results showed that under the condition of very low slurry pressure continuous injection of bentonite slurry, the slurry and pipe can also form a lubrication interface with a friction coefficient of about 10% of the friction coefficient of the pipe and soil contact interface.Chen [6] and Huang [7] determined the friction coefficient of the interface between silt, sand and steel pipe jacking, with or without lubricant slurry and under different lubricant slurry formulation conditions.The friction coefficients of the silt, sand and steel plate under the conditions of different ratios of lubricant slurry were compared to obtain the optimal lubricant-slurry matching ratio.
Peng et al. [8] carried out a study on the mechanical properties of the coarse-grained soil-concrete interface under different slurry conditions.It was demonstrated that the stress-strain curve was hyperbolic when the bentonite slurry was present on the contact surface.When the mixed soil slurry was present on the contact surface, the stress-strain curve had an obvious strain-softening phase, and the higher the cement content, the higher the peak strength.Zhang et al. [9] found that the water content of the slurry changed the shear stress-shear displacement curve of the interface, and when the water content was low, the curve was elastic-plastic in character, while when the water content was high, the curve was rigid-plastic in character.Zhao et al. [10] used the DIC technique to study the mechanical properties and mesoscopic failure mechanism of the gravel-sand-concrete interface with different relative densities.Yang et al. [11] used a ring shear apparatus to study the mechanical properties of the clay-concrete interface, with and without slurry lubrication.It was obtained that the presence of slurry lubrication caused the stressstrain curve to change from strain hardening to strain softening, and the strain softening characteristics were enhanced with the increase of normal stress.Zhang et al. [12] carried out a study on the mechanical properties of the pebble soil-concrete and uniform gravelsteel plate interface, with and without slurry lubrication conditions.It was concluded that the particle size, particle gradation and interface roughness of the soil were responsible for the inconsistent shear property patterns at the two different contact interfaces.Li et al. [13] used a direct shear test to study the friction characteristics between the contact surface of sand and concrete pipe jacking with different particle sizes.The test results showed that with the increase of the sand particle size, the friction resistance between the pipe and soil decreases, and mud can effectively reduce the friction resistance between the pipe and soil.Chen [14] investigated the mechanical properties of the interface of sandy soil and structures under different conditions of mud cake thickness and roughness of the structural interface.It was shown that when the interface was smooth, the stress-strain curve changed from a softening type to a mild softening type and finally to a double-folded type as the thickness of the mud cake increased.Li et al. [15] designed six different slurry ratios to study the interaction mechanism between a slurry-soil mixture and pipeline, and after pipe jacking grouting, analyzed the shear stress-displacement curve of slurry-soil concrete and the concrete interface under different conditions, and obtained the formula for calculating the shear strength of the slurry-concrete interface considering slurry content.
In summary, the current domestic and foreign research on steel pipe jacking wall friction resistance is mostly focused on the bentonite slurry formulation and the establishment and improvement of the pipe soil-pipe slurry contact model.There is little research on the calculation and influencing factors of the friction coefficient between the steel pipe section and the formation and the lubricating slurry.Most of the existing research is also on single specific working conditions, which is lack of representation.Therefore, the friction characteristics of sand and steel pipe jacking with different particle sizes under different normal stresses and different lubrication conditions are studied and analyzed by using the straight shear tester.The shear strength and residual strength of the interface of sand and steel pipe jacking measured by the test are regressed and analyzed, and the friction coefficient of the pipe and soil under different conditions is obtained according to the regression equation.

Test Plan 2.1. Testing Equipment
The test used a TSY-12D straight shear apparatus, which is shown in Figure 1.Additional settings and items used were the instrument maximum vertical load of 20 kN, the maximum horizontal thrust of 10 kN, the maximum horizontal shear displacement of 150 mm, straight shear apparatus on the shear box using organic glass material, the internal dimensions of 109 mm × 88 mm × 100 mm, the lower shear box size of 400 × 300 × 30 mm and pressurized end caps for 108 mm × 87 mm × 5 mm of stainless steel plate.During the test, the steel plate is placed in the lower shear box, and the upper shear box is filled with sand samples.The vertical servo motor applies constant normal pressure, while the horizontal servo motor pushes the lower shear box, so that the relative motion between the upper and lower shear box occurs at a constant rate, and the pressure sensor is used to measure the interface shear force.
sand and steel pipe jacking measured by the test are regressed and analyzed, and the friction coefficient of the pipe and soil under different conditions is obtained according to the regression equation.

Testing Equipment
The test used a TSY-12D straight shear apparatus, which is shown in Figure 1.Additional settings and items used were the instrument maximum vertical load of 20 kN, the maximum horizontal thrust of 10 kN, the maximum horizontal shear displacement of 150 mm, straight shear apparatus on the shear box using organic glass material, the internal dimensions of 109 mm × 88 mm × 100 mm, the lower shear box size of 400 × 300 × 30 mm and pressurized end caps for 108 mm × 87 mm × 5 mm of stainless steel plate.During the test, the steel plate is placed in the lower shear box, and the upper shear box is filled with sand samples.The vertical servo motor applies constant normal pressure, while the horizontal servo motor pushes the lower shear box, so that the relative motion between the upper and lower shear box occurs at a constant rate, and the pressure sensor is used to measure the interface shear force.

Steel Plate Preparation
The test is simplified for the steel pipe jacking section.The pipe section is simplified to the same material steel plate in order to ensure that the interface between the outer wall and the sand layer in the process of steel pipe jacking is consistent.The steel plate material used in the test is Q355, and the size is 350 mm × 250 mm × 25 mm.

Sand Preparation
The sand soil used for the test is self-assembled sand in the laboratory.The dried standard quartz sand is sieved into different particle sizes and then prepared according to the grading requirements of different sand soils in the Code for the Design of Building Foundations (GB 50007-2011) [16].The particle gradation of the different grain-size sand soil is shown in Figure 2.

Testing Material 2.2.1. Steel Plate Preparation
The test is simplified for the steel pipe jacking section.The pipe section is simplified to the same material steel plate in order to ensure that the interface between the outer wall and the sand layer in the process of steel pipe jacking is consistent.The steel plate material used in the test is Q355, and the size is 350 mm × 250 mm × 25 mm.

Sand Preparation
The sand soil used for the test is self-assembled sand in the laboratory.The dried standard quartz sand is sieved into different particle sizes and then prepared according to the grading requirements of different sand soils in the Code for the Design of Building Foundations (GB 50007-2011) [16].The particle gradation of the different grain-size sand soil is shown in Figure 2.

Lubrication Slurry Preparation
In the previous phase, the slurry is configured with different formulations and tested for basic performance parameters through indoor tests.The influence of the bentonite content, treatment agent type and dosage on the basic performance parameters of the slurry is analyzed to obtain the suitable dosage range of treatment agents in slurries of different concentrations.With reference to the results of the previous experiments, the slurry used for the test is bentonite slurry, and the slurry formula is 10% nano-bentonite

Lubrication Slurry Preparation
In the previous phase, the slurry is configured with different formulations and tested for basic performance parameters through indoor tests.The influence of the bentonite content, treatment agent type and dosage on the basic performance parameters of the slurry is analyzed to obtain the suitable dosage range of treatment agents in slurries of different concentrations.With reference to the results of the previous experiments, the slurry used for the test is bentonite slurry, and the slurry formula is 10% nano-bentonite + 0.3% high viscosity sodium carboxymethyl cellulose + 0.3% xanthan gum.The lubricated slurry performance parameters are as follows: plastic viscosity 20.0 mPa•s, dynamic cutting force 24.5 Pa, initial cutting force 14.3 Pa, final cutting force 15.8 Pa, funnel viscosity 150 s, and filter loss 6.4 mL•30 min −1 .

Test Steps and Precautions
The test procedure and precautions are as follows.
(1) Installation of steel plate.Place the steel plate smoothly in the lower shear box and fix it.
Pour in the slurry.Pour the configured slurry evenly onto the surface of the steel plate (no slurry lubrication control group without this step).(3) Fill and compact the specimen soil body.According to the density of the different particle-size gradations of the specimen, the soil body is calculated and accurately weighed and loaded into the upper shear box, poured evenly with a wooden mallet and compacted, then covered with a pressurized cover.(4) Apply normal stress set horizontal stress.By controlling the vertical servo motor to adjust the vertical loading screw down and touch the pressurized cover, after the test soil is stable, adjust the normal stress to 80 kPa, 60 kPa and 40 kPa.Then, control the horizontal servo motor and adjust the horizontal loading screw to push the lower shear box at a constant rate.Stop shearing and record the data when the horizontal pressure sensor is stable.(5) After the test, the load is slowly removed, the sand in the upper shear box is taken out and the steel plate in the lower shear box is cleaned to carry out the next set of tests.(6) In the slurry lubrication test group, the sand is filled in the upper shear box and the slurry is filled.After scraping, the subsequent test steps were repeated to continue the test.(7) Before the test, the shear stress of the empty shear box should be tested firstly, and the test shear stress should be subtracted from the initial shear stress during the test process.Each group of tests is repeated three times, and the test results are averaged.
If the test results are quite different, the test is repeated, and the three groups of average values close to the test results are taken as the test results.

Effect of Particle Size on Shear Stress
It can be seen from Figure 3 that the variation of the shear stress-displacement curve of sandy soil, with and without slurry lubrication, is basically the same at a normal stress of 40 kPa.In the initial stage of shear, with the increase of shear displacement, the shear stress increases, and there is an obvious peak value of the curve.After reaching the peak, the shear stress will decrease with the increase of the shear displacement and then gradually stabilize, showing an obvious strain-softening type.Figure 3a shows that without slurry lubrication, the peak and stable values of the shear stress are related to the grade of the soil sample, which shows that powder sand > fine sand > medium sand > coarse sand > gravel sand, that is, the peak and stable values of the shear stress decrease with the increase of the sand soil particles.Figure 3b shows that the order of peak shear stress and the stable value in the slurry-lubricated test group is the same as that without slurry lubrication, but the value is smaller than that without slurry lubrication.The following test results can be obtained by comparing the test groups with reference to different normal stress conditions: the larger the sand particles are, the smaller the shear stress between the steel plate and the sand, when other conditions are the same.Sitharam et al. [17] in 2000 showed that the internal friction angle of sandy soil is positively related to its average grain size, and as the grain size of sandy soil increases, the internal friction angle of sandy soil increases, and the shear stress between sandy soil and a steel plate increases.The above experimental phenomenon is just the opposite of Sitharam, which is analyzed as follows.There are differences in the mechanical parameters between the contact surface of sandy soil and a steel plate with different grain sizes, which are related to the sandy soil grain size and grain gradation.The sandy soil configured in the test is divided into five types according to the average particle size.The sandy soil particles are all irregularly shaped with sharp angles, and the sandy soil with a large average particle size has a high content of large particles.The contact mode between large particles is mostly contact, the pores between particles are large and the effective contact area is relatively small, so the effective contact area between the same steel plate is small.The sand with a small average particle size has a high content of small particles, a good filling effect between particles and a denser structure, so the effective contact area with the steel plate is also larger.Therefore, the peak shear stress between the sand with a small average particle size and the steel plate is larger under the same condition of normal stress.
above experimental phenomenon is just the opposite of Sitharam, which is analyzed as follows.There are differences in the mechanical parameters between the contact surface of sandy soil and a steel plate with different grain sizes, which are related to the sandy soil grain size and grain gradation.The sandy soil configured in the test is divided into five types according to the average particle size.The sandy soil particles are all irregularly shaped with sharp angles, and the sandy soil with a large average particle size has a high content of large particles.The contact mode between large particles is mostly point contact, the pores between particles are large and the effective contact area is relatively small, so the effective contact area between the same steel plate is small.The sand with a small average particle size has a high content of small particles, a good filling effect between particles and a denser structure, so the effective contact area with the steel plate is also larger.Therefore, the peak shear stress between the sand with a small average particle size and the steel plate is larger under the same condition of normal stress.

Effect of Normal Stress on Shear Stress
The shear stress-displacement curves of each group in Figure 4 were measured under the condition of non-lubricated slurry and the analysis of the shear stress-displacement curves of sand samples with different particle sizes under the normal stress of 40 kPa, 60 kPa and 80 kPa.It can be concluded that the peak and stable values of the shear stress between the contact surface of the sand and steel plate increase as the normal stress increases.The reason is that when the normal stress increases, the sand and soil compactness increases, the porosity decreases, the extrusion between the soil particles is more intense, the friction increases, and the shear stress between the contact surfaces increases as a result.With the increase of normal stress, the shear displacement corresponding to the maximum shear stress in the sandy soil increases.The reason is that for the same particle gradation sand sample, when the normal stress increases, the interaction force between the particles increases.During the shear process, the frictional force between the particles increases, and the relative position between the soil particles is difficult to change and cannot reach a relative stable state in a short time [4].The strain-softening phenomenon occurs in sandy soils under different normal stresses, and the shear stress-reduction value increases with the increase of normal stress, but it has no effect on the trend of the normal stress-displacement curve.The reason may be that normal stress only changes the magnitude of the interaction force between the sand and soil particles, without changing their

Effect of Normal Stress on Shear Stress
The shear stress-displacement curves of each group in Figure 4 were measured under the condition of non-lubricated slurry and the analysis of the shear stress-displacement curves of sand samples with different particle sizes under the normal stress of 40 kPa, 60 kPa and 80 kPa.It can be concluded that the peak and stable values of the shear stress between the contact surface of the sand and steel plate increase as the normal stress increases.The reason is that when the normal stress increases, the sand and soil compactness increases, the porosity decreases, the extrusion between the soil particles is more intense, the friction increases, and the shear stress between the contact surfaces increases as a result.With the increase of normal stress, the shear displacement corresponding to the maximum shear stress in the sandy soil increases.The reason is that for the same particle gradation sand sample, when the normal stress increases, the interaction force between the particles increases.During the shear process, the frictional force between the particles increases, and the relative position between the soil particles is difficult to change and cannot reach a relative stable state in a short time [4].The strain-softening phenomenon occurs in sandy soils under different normal stresses, and the shear stress-reduction value increases with the increase of normal stress, but it has no effect on the trend of the normal stress-displacement curve.The reason may be that normal stress only changes the magnitude of the interaction force between the sand and soil particles, without changing their contact mode and movement, and the shear stress-strain curve only changes the numerical magnitude, without changing the curve direction and trend.When the normal stress is larger, the interaction force between the sandy soil particles is larger, the soil particles transport violently, the density of soil particles near the contact surface of the sandy soil and steel plate decreases, and the strain softening is more obvious at this time.
contact mode and movement, and the shear stress-strain curve only changes the numerical magnitude, without changing the curve direction and trend.When the normal stress is larger, the interaction force between the sandy soil particles is larger, the soil transport violently, the density of soil particles near the contact surface of the sandy soil and steel plate decreases, and the strain softening is more obvious at this time.

Effect of Lubrication State on Shear Stress
Under the condition of normal stress of 60 kPa, the peak shear stress and stability of sand with different particle sizes under different lubrication conditions are shown in Table 1.In the analysis obtained, the peak shear stress in each group with slurry lubrication is about 74% of that without slurry lubrication, and the stable value of shear stress is about 70% of that of the group without slurry lubrication.Considering continuous grouting in practical engineering, a complete slurry sleeve can be formed and the permeability of the stratum is smaller than that of the sand.The test results show that injecting lubricating slurry between the soil and steel pipe jacking during steel pipe jacking construction can effectively reduce the friction between the pipe and soil.

Effect of Lubrication State on Shear Stress
Under the condition of normal stress of 60 kPa, the peak shear stress and stability of sand with different particle sizes under different lubrication conditions are shown in Table 1.In the analysis obtained, the peak shear stress in each group with slurry lubrication is about 74% of that without slurry lubrication, and the stable value of shear stress is about 70% of that of the group without slurry lubrication.Considering continuous grouting in practical engineering, a complete slurry sleeve can be formed and the permeability of the stratum is smaller than that of the sand.The test results show that injecting lubricating slurry between the soil and steel pipe jacking during steel pipe jacking construction can effectively reduce the friction between the pipe and soil.

Calculation of Mechanical Parameters of Contact Surface
Regression analysis of the shear strength τ p (the peak shear stress in the τ-s curve of the interface) and residual strength τ r (the stable value of the shear stress in the τ-s curve of the interface) at the contact surface between sandy soil of different particle sizes and steel plates, with and without slurry, under different normal stresses was completed.To analyze the effect of the sand particle size on mechanical parameters, such as the interface internal friction angle ϕ, interface cohesion c and interface sliding friction coefficient µ, the interface internal friction angle ϕ and interface cohesion c are regressed according to Equation (1), and the interface sliding friction coefficient µ is regressed according to Equation (2).The shear strength τ p and residual strength τ r of the interface between sandy soil of different particle sizes and steel plates, with and without slurry, are shown in Table 2, and the regression equations of the shear strength τ p and residual strength τ r are shown in Tables 3 and 4, respectively.
where σ-normal stress, kPa; τ p -interface shear strength, kPa; τ r residual strength, kPa; ϕ-interface internal friction angle, • ; c-interface cohesion, kPa; and µinterface sliding friction coefficient.The regression equations of shear strength τ p and residual strength τ r at the interface between the sand and steel plate, with and without slurry, are shown in Tables 3 and 4, respectively.
Combined with the slope and intercept of the regression equation in Tables 3 and 4, we calculated the interface internal friction angle ϕ, the interface cohesion c and the interface sliding friction coefficient µ, as shown in Table 5.From the data in Table 5, it can be obtained that in terms of the interface sliding friction coefficient, µ powder sand > µ fine sand > µ coarse sand > µ medium sand > µ gravelly sand ; in terms of the interface internal friction angle, ϕ powder sand > ϕ fine sand > ϕ coarse sand > ϕ gravelly sand > ϕ medium sand ; and in terms of the interface cohesion, c medium sand > c powder sand > c fine sand > c coarse sand > c gravelly sand .
There are differences in the mechanical parameters of the interface between the sand and the steel plate with different particle sizes, which are related to factors such as the sand size, particle gradation and roughness of the pipe itself.In terms of the interface sliding friction coefficient, the friction coefficient of the interface between the sand with smaller particle size and the steel plate is generally greater than that between the sand with larger particle size and the steel plate.In terms of the internal friction angle, the internal friction angle of the interface of the sand and steel pipe with smaller particle size is larger than the internal friction angle of the interface of the sand and steel pipe with larger particle size, while the law is not obvious in the interface cohesion.The reason for the above phenomenon is that the coarse particle content of larger-size sandy soil is greater, the pores between the particles are also larger, the fine particle content is less, and the pores are difficult to completely fill, so the effective interface area between them and the pipe is smaller.The sand with smaller particle size is denser due to a better pore filling effect, so the effective area between the sand and pipe is larger.Therefore, under the same normal stress condition, the shear strength and residual strength of the interface of the sand and pipe are larger, so the interface internal friction angle and sliding friction coefficient are larger.

Conclusions
(1) The shear stress of sandy soil and steel plate is influenced by the particle size, gradation and compactness.It increases with the decrease of the average particle size of the sandy soil.The main reason is that the effective contact area between the sandy soil with small average particle size and the steel plate is large, and the interaction force between the soil particles and the steel plate is also larger.(2) The magnitude of normal stress has no significant effect on the shear stress-displacement curve between the sand and the steel plate, but it has an effect on the peak shear stress, the stability value and the shear displacement corresponding to the peak on the shear stress-shear displacement curve at the interface.Among them, the peak shear stress and the stability value are positively correlated with the normal stress, while the shear displacement corresponding to the peak shear stress increases roughly with the increase of the normal stress.(3) With or without slurry lubrication, the shear stress curves between the sand and steel plate are a strain-softening type, and the use of slurry lubrication can effectively reduce the shear stress between the sand and steel plate by 20-30%.In practical engineering, the lubrication effect will be better due to the continuous grouting and the different permeability of the formation.(4) When there is no lubricated slurry, the mechanical parameters of the interface between the sand and steel plate are related to the sand particle size, particle gradation and roughness of the pipe itself.The friction coefficient and internal friction angle of the interface of the sandy soil and steel plate decreased with increasing particle size.The friction coefficient ranged from 0.45-0.54, the internal friction angle ranged from 26-29 • , and the cohesive force of the interface ranged from 5.41-6.72 kPa.

Figure 4 .
Figure 4. Shear stress-shear displacement curves of different normal stress.

Figure 4 .
Figure 4. Shear stress-shear displacement curves of different normal stress.

Table 1 .
Shear stress peak and stable value of sand interface under different lubrication conditions.

Table 1 .
Shear stress peak and stable value of sand interface under different lubrication conditions.

Table 2 .
Data of shear strength and residual strength of sand-steel interface with different particle sizes.

Table 3 .
Regression equation of shear strength and normal stress of interface.

Table 4 .
Regression equation of residual strength and normal stress of interface.

Table 5 .
Mechanical parameters of interface between sand with different particle sizes and steel plate.