Abstract
The application range of the earth pressure balance (EPB) shield has been expanded due to advances in optimization methods, one of which is the application of foam conditioning. This method is widely used in EPB tunnelling owing to its strong applicability in different hydrological and geological conditions. When applying the foam conditioning method under different circumstances, it is necessary to optimize the conditioning parameters. Therefore, it is necessary to propose a test procedure for evaluating foam properties and the conditioning effect. This paper proposes a procedure to assess foaming agents by the mixing test and microscopic observation of the foam and a procedure that combines the slump test, compression test, and shear test to assess the foam-conditioned soil and determine the optimal parameters of conditioning. The test method is introduced and performed on fully weathered granite from Guangzhou Metro Line 21. The test results demonstrate that the foam injection ratio and pressure and type of the foaming agent all influence the performance of the conditioned soil. Moreover, the suggested conditioning scheme is proposed, and the application of the scheme can improve the tunnelling efficiency.
1. Introduction
The EPB shield is being more and more widely used in the construction of tunnels, and many related studies have been done on this technique [1–7], in which more situations and construction technology are considered. The adaptability of EPB to challenging geological conditions is getting better, in part because of the extensive application of soil conditioning. Soil conditioning, which is achieved by injecting conditioning agents into the soil of the tunnel face and excavation chamber, is used to guarantee effective EPB shield performance and to reduce clogging and cutterhead wearing [8–13]. Conditioning agents include foam, polymer, bentonite, and dispersant. Each conditioning agent has its own application scope, depending on the hydrological and geological conditions and the type of the shield machine. Foaming agent is the most widely used agent due to its strong applicability in various conditions. Many researchers use different methods to quantify the adhesion properties of different soils and the improvement of foaming agents. Galli and Thewes [1] and Thewes et al. [10] discussed the methods to determine the quality of foams and soil-foam mixtures and pointed out that the foaming agent type mattered in soil conditioning. Some researchers proposed methodologies [14–16] to assess the clogging potential in different situations; in these classification systems, the chart proposed and revised by Thewes and Holloman [13, 17] is widely accepted. Langmaack [18, 19] studied the characterization of foams and polymers and discussed some parameters of conditioning foam, such as concentration of foaming solution, foam expansion ratio [15], and foam injection ratio (FIR). Liu et al. [20] investigated the change of Atterberg limits to evaluate conditioning methods for the EPB shield, the Atterberg limits decrease as the foam was injected, and the clay becomes more easily dispersed. de Oliveira et al. [21] reported the application of the consistency index to assess the material to be excavated by the EPB machine. Bezuijen [22] studied the impact of soil permeability on the properties of a foam mixture to improve soil conditions because the foam will increase the porosity between grains, reducing the permeability to avoid spewing failure. Peila [8], Peila et al. [23, 24], and Zumsteg and Puzrin [25] conducted laboratory tests to develop the conditioned clay soil and rock mass assessment, proposed a new approach and devices to measure the adhesion of soil, and demonstrated that the injection of polymer can improve the mechanical properties of the conditioned soil. Zumsteg and Puzrin [25] used shear plate apparatus to quantify the stickiness and adhesion of conditioned clay soil samples and proposed new quantitative assessment methods, using shear plate apparatus to measure the adhesion between soil and steel. Budach and Thewes [26] discussed the feasibility of soil conditioning in coarse-grained soil and developed the EPB shield application ranges in coarse ground. Mori et al. [27, 28] assessed conditioned clay soil and sand behaviors under pressure with a pressurized test chamber and demonstrated that the performance of the conditioned soil was related to the pressure and foam parameters. Vinai et al. [11] used the slump test and a screw conveyor device to define the characteristics of conditioned sand. Ye et al. [29] found the optimum conditioning parameters in argillaceous siltstone and carried out field implementation, finding that soil conditioning with the foaming agent can efficiently reduce adhesion and clogging on site. However, most methods of evaluating the conditioned soil only focus on one aspect of the physical and mechanical properties of the conditioned soil, or the test apparatus is complicated and not easy to apply on the construction site. In addition, most of the test conditions do not take the pressurizing of conditioned soil into consideration.
This paper presents a scheme that includes a foam mixing test, slump test, compression test, and shear test to study the performance of foaming agents and the properties of conditioned soil and determine the optimal parameters of soil conditioning in specific conditions. The procedure proposed in this paper aims to characterize the effects of soil conditioning with a relatively simple method which can be applied in practice. Based on the results of the test, the soil conditioning parameters for fully weathered granite from Guangzhou Metro Line 21 are proposed, and the application of the conditioning scheme on the construction site is discussed.
2. Engineering Conditions
The plan view of the left line of the Guangzhou Metro Line 21 Jinkeng–Zhenlongnan section is shown in Figure 1. It is constructed by an EPB shield with an external diameter of 6.0 m. The length of this section from the No. 1 air shaft to the No. 2 air shaft is more than 3 km, and the buried depth of the tunnel is 12.00∼22.0 m. The urban development along the section is extremely dense and old. The buildings often have low foundations, which are susceptible to deformation caused by shield construction. The tunnel mainly passes through strongly weathered and completely weathered granite. The surface water is abundant, with reservoirs and numerous small rivers along the section, the groundwater contains pore water and fissure water, and the average stable water level is 3.78 m. During the construction of this section, serious clogging occurred due to the high content of cohesive particles. Clogging will limit the pressure control accuracy, the performance of the cutting wheel, and advance speed and will increase the disturbance of the tunnel face. The surface settlement will enlarge and needs to be strictly controlled in this section. Soil conditioning was performed in order to reduce clogging. However, the parameters of soil conditioning on the job site are determined empirically, and the rationality of the conditioning parameters has not been verified.
3. Foaming Agents Properties Assessment
Four foaming agents widely used in China were tested in the mixing test. In order to find out the optimum concentration of the foaming agents, the test groups of each foaming agent were carried out with the concentration of 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, and 9%, respectively.
Foaming ability and stability of the foam are two main factors for the evaluation of foaming agents [30, 31]. The performance of the foaming agents was investigated with the Waring blender method [32, 33]. This method is simpler and easier in practical operation than other methods for evaluating the performance of the foaming agent, and thus, the test can be performed on the construction site. The sketch of this test is illustrated in Figure 2. In the test, 200 ml liquid, a mixture of water and surfactant in a certain proportion, was prepared in a beaker with a volume of 1000 ml. The liquid was mixed for 3 min at 2000 rpm. Foaming ability and stability of the foam were measured by the volume and half-life of the foam, respectively. The volume of foam was recorded at an interval of 30 s, and the half-life is regarded as the duration that the volume of foam decays into half the volume, that is, the time 100 ml liquid separated out. In addition, the foam structure is observed to evaluate the microscopic properties of the foam.
3.1. Foaming Ability of Agents
Foaming ability refers to the ability of a certain amount of the foaming agent to produce foam, and this indicator is assessed by the volume of foam in the mixing test. The test results of foaming agents at different concentrations are presented in Figure 3(a). The results indicate that when the concentration of liquid is below 3%, foam volume increases significantly with the increase of concentration, which results from the surface viscosity and surface elasticity increment of the agent solution. When the concentration is higher than 3%, the growth of foaming volume slows down, and the surfactant concentration in the liquid is close to the critical micelle concentration (CMC). Furthermore, the surface tension of the solution is lowered; thus, the energy required for foaming is lowered, and the foaming ability of the solution is improved. The foaming property of the agent solution will not be enhanced significantly if the concentration is higher than 3%; thus, the increase of foam volume slows down.
(a)
(b)
Among the four foaming agents, foaming agent B has the largest increase at low concentration, but the foaming volume at a concentration of 0.5% is only 720 ml, and the maximum foaming volume is 875 ml, which indicates that the foaming effect is poor and the agent would be fully exerted at a certain concentration. Foaming agents A, C, and D have similar foaming volume curves. Foaming agent C has the smallest increase at low concentration and reaches the maximum foaming amount at a concentration of 2%. The foaming volume of the concentration of 0.5% is 925 ml, and the maximum foaming volume, which is at the concentration of 4%, is 975 ml, indicating that the foaming effect is good and the agent is effective at low concentrations.
3.2. Half-Life of Foam
The stability of the foam is evaluated by its half-life. The results are shown in Figure 3(b). As can be seen, the half-life increases strongly at concentrations below 3%. When the concentration is higher than 3%, the half-life increases at an extremely slow speed. The increase of surface viscosity makes the drainage of the foam slower. In addition, the thickness of the liquid film of the foam will be strengthened, which will make the foam stable. Like the volume of foam, the strength of the liquid film reaches a bottleneck after the concentration reaches 3%. Given the volume and half-life of foam and taking engineering economics into consideration, it is reasonable to use a foam concentration of 3%.
Foaming agent B is obviously superior to the other foaming agents, with a half-life of 4.2 min at a concentration of 0.5%. However, during the construction process, the excessively long half-life value of the foam, which indicates that the foam takes a long time to dissipate, will make it more difficult to improve the dumped soil treatment. Although foaming agents A and D have higher foaming amounts, their half-life is shorter. Moreover, the foam is in a complicated stress environment during construction; thus, the performance of the foam with low half-life will be limited in practice.
3.3. Foaming Speed
Foaming speed is also a crucial factor in practical construction. The foam is prepared by mixing the solution with air in the foam gun at the construction site [19]. Since the length of the foam gun is limited, the mixing time at the site is limited, which indicates that the foaming speed is also an important indicator of foaming agent quality. The volume of the foam at different mixing times is shown in Figure 4. It can be seen that the foaming speed of agent C is the fastest among the 4 foaming agents; the foaming is complete at about 1.5 minutes. The foaming speed of the other three foaming agents is similar, about 2 minutes to fully foam. Agent C can fully foam with less energy in a short period of time, which can enhance the adaptability of foam conditioning to the actual situation.
3.4. Microscopic Investigation of the Bubble Size
Microscopic property is the fundamental physical characteristics of foam and is related to the stability of foam [34]. Thus, microscopic observation of the foam is carried out in this paper as an add on to foam property evaluation. The foam observation system is demonstrated in Figure 5. The microscope above the foam container can be used to observe the foam, and the viewing range and location of the microscope is fixed. The image of the stable foam can be captured by the computer linked to the microscope. The foam is generated by the foaming system according to the required parameters, and then, the foam flows into the container via the inlet. When the foam condition and flow rate are stable, close the inlet and outlet valves and capture the image of the foam immediately. Based on the images of the foam with different parameters, numbers of foam in different diameter ranges are counted, and the foam bubble size and size distribution could be analyzed. Taking agents C and D, for example, foam generated by the 3% concentration of agent solution with the FER of 10 is investigated in the test, and the foam bubble diameter distribution is shown in Figure 6.
As shown in the picture above, the foam of agent C has a smaller diameter (<0.2 mm) foam bubble, which accounts for 80% of the total bubble amount. Smaller foam size may result in more stable foam [10]. In addition, the distribution of bubble size amount is fitted with a cubic function in the picture, which indicates that the distribution of agent C’s foam bubble size is more concentrated, and the bubble size is more uniform. As for agent D, there are some relatively big foam bubbles, and the distribution is more dispersed. This paper proposes the foam observation method as one of the test methods for evaluating foam properties.
4. Assessment of Conditioned Soil
4.1. Preparation of Conditioned Soil
The investigated soil was collected from the No. 21 line of the Guangzhou Metro at the location marked in Figure 1. The fully weathered granite was brownish yellow. The original rock structure had been basically destroyed. The soil is mainly composed of powdery clay, which is easily pinched by hand. It softens and disintegrates easily when exposed to water, and the water permeability is low. The geotechnical parameters of the soil are shown in Table 1.
Before the test starts, the investigated soil was dried, and grains larger than 10 mm was removed to ensure the conditioning effect [24, 35] and the reliability of the test. In the test, the soil was firstly mixed with water to reshape the soil with a water content of 22%, which is the same as that of the place where soil was collected. And in the soil conditioning test, a certain volume of foam was mixed uniformly with the reshaped soil. The FIR is the volume of the injected foam compared to the volume of the soil. The particle size grading curve of the soil used in the five groups of tests is shown in Figure 7.
The self-developed foaming system is shown in Figure 8. In the tests of conditioned soil, the foaming liquid was a 3% concentration liquid of foaming agents C and D, respectively. The foaming pressure was 0.3 MPa, and the FER in this interval was set to 10, which is the same as the foaming parameters of the construction site.
During the preparation process, 3 kg of a 3% concentration of conditioning solution was placed in the liquid storage tank, the air compressor was opened after the liquid storage tank was completely sealed, and the air pressure was adjusted to 0.3 MPa by the air control valve. The gas and liquid flow rates were adjusted to 2 L/min and 200 mL/min, respectively, by the flowmeter. When the foam flow was continuously stabilized, the foam could be mixed with the soil using a mixer. The conditioned soil thus obtained will be used for the slump test, compression test, and shear test.
4.2. Slump Test
The slump is a common indicator for characterizing conditioned soil properties [9, 24, 29, 36, 37], and the slump test is widely used due to its low cost and simple operation. The slump test is an indicator of the workability of soil and provides an operational evaluation of soil conditioning methods. Three groups of parallel tests were conducted under each foam injection ratio. Several slump tests were conducted under different injection ratios of agents C and D. Images of the slump tests of agent C at various injection ratios are shown in Figure 9, and the test results are demonstrated in Figure 10.
(a)
(b)
(c)
(d)
(e)
As the images show, not only does the slump increase with the injection ratio but the uniformity of the conditioned soil also increases. At the low injection ratio, the foam is easily drained from the soil and cannot exert the dispersion effect. As the injection ratio increases above 40%, the foam’s existential environment is improved, and the foam begins to function effectively as a dispersing clay particle, smoothing the surface of the soil sample. It can be seen from Figure 10 that the slump value increases with the injection ratio, and the growth slows after more than 60%. When the foam injection ratio exceeds 60%, the amount of foam exceeds the capacity of the soil. In addition, an injection ratio exceeding 80% is so large that the economic efficiency of the soil conditioning scheme will decrease.
The results of the slump tests demonstrate that the improvement effect of product C is better than that of product D. The slump of the soil conditioned by agent D can reach more than 10 cm when the injection ratio is greater than 60%. The optimal range of slump for conditioned soil varies in different researchers [29, 38–40]. But a large number of scholars [10, 37, 40] suggested that the slump value of 10∼15 cm could meet the EPB shield requirement. The test results revealed that the FIR needed to be around 60% to meet the fluidity requirement that the slump exceeds 10 cm.
4.3. Compression Test
Appropriately conditioned soil has good compressibility, which enables it to form a plastic zone with gradually weakening stiffness at the face of the shield machine. It can play a buffering role in the excavation process and solve the problem that the shield cutter disk is not sensitive to the soil reaction. At the same time, it also helps to maintain the stability of the face.
The compression test device is shown in Figure 11(a). The sample soil is placed in a cylinder with an inner diameter of 20 cm and a height of 30 cm. The bottom and side walls of the cylinder are fixed by four screws. The cover plate of the steel drum is 3 cm thick, which is the pressure boundary. The pressure is applied by a hydraulic jack, and the pressure value is indicated by the hydraulic gauge at the top. Displacement of the cover plate under different pressures is measured with a digital Vernier caliper. In this test, the maximum stroke of the cylinder of the hydraulic jack was 18 cm. The range of the hydraulic gauge was 0 MPa∼1.2 MPa, and the precision was 0.02 MPa. The range of the digital Vernier caliper was 0 cm∼15 cm, and the precision was 0.01 mm.
(a)
(b)
In the test, the conditioned soil was divided into five batches and placed in the cylinder in layers, and then the cover was placed horizontally. After that, the hydraulic jack was pressurized at a constant speed, and the displacement of the cover plate was recorded at intervals of 0.05 MPa from 0 to 0.5 MPa. The test was carried out at injection ratios of 0%, 20%, 40%, 60%, and 80%, respectively, and three parallel tests were performed at each injection ratio. The test groups are shown in Table 2.
The mean values of the data from the three tests are shown in Figure 12.
(a)
(b)
As can be seen in Figure 12, the compression improvement effects of agents C and D are similar when the FIR is lower than 40%; the maximum compression amount at 40% injection ratio is about 40 mm. The compressive amount of soil under each injection ratio keeps increasing when the pressure is lower than 0.3 MPa. After that, improvement effects of agents C and D are improved as the FIR increases, and the gap between the curves of each injection ratio gets larger and larger at the same time. The maximum compression (at 0.5 MPa) of agent D is 47.45 cm and that of agent C is 68.55 cm. In the early stage of the compression process, the increase of the compressive amount includes the decrease of the soil particle gap, foam volume, and drainage. In the later stage, the foam in the soil gradually ruptures under high pressure, and the increase of the compression amount only includes the reduction of the soil particle gap and drainage. If the pressure continues to grow, the soil is compacted to form an overconsolidated soil, which is prone to result in clogging.
For agent C, when the injection ratio is below 80%, the compressibility of the soil increases with the increase of the injection ratio, but the increase of the improvement effect is significantly reduced after the injection ratio exceeds 60%. However, after the injection ratio reaches 60%, the soil compression still has an upward trend when the pressure reaches 0.5 MPa, which indicates that the foam of agent C remains stable under high pressure. Compared with agent D, agent C has a continuous conditioning effect under high pressure when the injection ratio reaches 60%. The difference between the two products reveals that the foam stability of agent C is better than that of agent D.
From the analysis above, and taking cost issues into consideration, the optimum injection ratio of agent C is 60%, the optimal earth pressure is 0∼0.5 MPa, and the pressure application range increases with the increase of injection ratio. The suggested injection ratio of agent D is 40%, and the approximate application pressure range is 0∼0.3 MPa; the conditioning effect in compression is worse than that of agent C.
4.4. Shear Test
The shear strength of the conditioned soil will directly affect the excavation of the EPB shield machine. Under the appropriate shear strength, the torque of the cutting wheel will be low; thus, the cutterhead temperature will be lowered, and the cutter tool wear and the risk of clogging will decrease. At the same time, the shield tunnelling process will be smoother, and the tunnelling efficiency will be considerably improved.
The shear test device demonstrated in Figure 11(b) is substantially the same as the compression test device. The difference is that a force-transmitting rod with a diameter of 2 cm is arranged, and the top of the force-transmitting rod is welded with four shear plates at an angle of 90 degrees. The dimension of the shear plate along the direction of the transmission rod is 7 cm, while the dimension perpendicular to the direction of the transmission rod is 9 cm, and the thickness of it is 0.2 cm. In the test, the top of the force-transmitting rod was controlled at a position 15 cm above the bottom of the cylinder. The bottom of the force-transmitting rod was connected with the torque meter. The torque measurement range was 0 N·m∼6 N·m, and the accuracy was 0.06 N·m. During the test, the bottom of the transfer bar was first driven out of the hole at the bottom of the pail and connected to the torque meter. After that, the investigated soil was filled into the cylinder in the same way as the compression test. After loading, the soil body was pressurized to the specified pressure, the torque meter was rotated with uniform speed, and the measurements were recorded. The test groups are shown in Table 3.
The average values of the three shear test results are shown in Figure 13.
(a)
(b)
As can be seen from Figure 13, both agents C and D can effectively reduce the torque, and agent C is superior to agent D in improving the soil under various pressures. The improvement effects of the two agents are greatly influenced by the pressure, and the improvement effects of agents C and D under atmospheric pressure (the curve of 0 MPa) are obviously better than that under additional pressure; that is, the torque decreases rapidly with increasing injection ratio. The injected foam can effectively disperse the soil particles and transform the soil into a plastic state, thus rapidly reducing the torque. When the injection ratio is over 40%, the increase of torque improvement decreases significantly because the foam has almost completely separated the soil particles.
When the injection ratio of agent C is 20%, take the average value of torque at different pressures into consideration, the torque reduction accounts for 38.94% of the total torque reduction (difference in torque average at 0% and 80% injection ratio). The reduction of the average torque for agent D is 14.11%. The gap between the two agents proves that the foam of agent C could have a better improvement effect at the initial stage of injection due to its stability. Furthermore, the torque improvement effects of both agents C and D exceed 95% when the injection ratio is 60%, which indicates that the suggested injection ratio is 60%.
4.5. Engineering Application
In the construction of the section between No. 1 air shaft and No. 2 air shaft of Guangzhou Metro Line 21, different conditioning schemes were used in sections. The suggested conditioning parameters, namely, the conditioning agent C, the surfactant concentration = 3.5%, and the FIR = 60%, were utilized from lining 944 to 1094 in order to ensure the safety of the project when crossing the area covered by aged buildings. In the previous conditioning scheme used from lining 770 to ring 943, agent D was used with the surfactant concentration = 3.5% and FIR = 40%.
On the construction site, the pressure in the soil chamber varied from 0.11 MPa to 0.15 MPa, which is in the proposed application range of pressure based on the compression test; the suggested range of thrust is 10 MN∼15 MN, and the suggested torque is up to 2.9 MN·m. The torque and thrust of the cutting wheel recorded from lining 770 to 1094 are demonstrated in Figure 14. Compared to the previous section, namely, lining 770∼943, both the torque and thrust of lining 944∼1094 have declined after adopting the new conditioning scheme. The average torque is reduced by 13.5% from 1.93 MN·m to 1.67 MN·m, while the average thrust decreased by 5.6% from 14.03 MN to 13.25 MN. Moreover, it was reported that the modified conditioning scheme can reduce the frequency of chamber opening as well as the risk of clogging.
(a)
(b)
As shown in Table 4, the standard deviations of torque and thrust have dropped significantly. The torque and thrust become more stable after the application of the new conditioning scheme, indicating that the workability of conditioned soil was enhanced, which makes the excavation smoother, saves energy resources, and enhances the controllability of EPB shield tunnelling.
5. Conclusions and Outlook
The test procedure proposed to assess foaming agents and conditioned soil is verified to be rational through several tests. In the performed tests of different foaming agents, both the foam volume and its half-life indicated that a foam concentration of about 3% was reasonable. From the overall situation of half-life, foaming amount, and foaming speed, agents C and D were better among the four foaming agents. The microscopic investigation of the bubble size can help to characterize the foam, and the adaptability of foam bubble size and the influencing factors of foam bubble size remain to be further studied in later research. For the conditioned soil tests, the results of the slump test, compression test, and shear test consistently demonstrated that an injection ratio of about 60% can effectively improve fluidity, compressibility, and shear strength and can ensure the economical efficiency of construction. In addition, the test results of the conditioned soil showed that agent C is more efficient than agent D in its conditioning effect. Therefore, the assessment procedure proposed in this paper can effectively evaluate the performance of a foaming agent and its conditioning effect on soil. Targeting the problems of adhesion and clogging in the clay layer represented by fully weathered granite, this paper carried out laboratory tests to determine the optimal soil conditioning method. The following conclusions can be made:(1)The performance of the foaming agent can be quickly and comprehensively verified by the mixing test. The foaming ability and stability of the tested foaming agents both increase with the increase of foaming surfactant concentration until 3%, and a concentration of 3%∼3.5% is suggested to be applied in construction sites.(2)The performance of conditioned soil can be assessed by the proposed scheme, which includes a slump test, compression test, and shear test. The tests demonstrate that foam conditioning can significantly improve the fluidity and compressibility and reduce the shear strength so that the workability of the soil can be significantly improved.(3)Among the agents tested in this paper, agents C and D are superior in foam quality, but agent C is better than D in the actual conditioning effect based on the test results of the conditioned soil; thus, agent C is recommended to be used on the construction site.(4)For the fine-grained fully weathered granite with 22% water content in the No. 21 line of the Guangzhou Metro, the optimum FIR ranges from 60% to 80%. And the application of the suggested conditioning parameters can improve the tunnelling efficiency, reducing and stabilizing the torque and thrust of the cutting wheel.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there are no conflicts of interest.
Acknowledgments
This research was funded by the National Key Research and Development Program of China (no. 2016YFC0802205), the National Natural Science Foundation of China (no. 51578460), and the Key Research and Development Program of the Sichuan Science and Technology Plan (no. 2017SZ0043).