Optimization and Field Validation of Soil Conditioning Scheme for EPB Shield Tunneling in Cobble–Boulder Stratum: Case Study on Beijing Metro Line 16

Abstract

As laboratory experiments evaluating soil strata conditioning schemes for earth pressure balance shield (EPBS) tunneling are limited by the size of the test equipment, large pebbles and boulders are typically replaced by an equal mass of smaller pebbles (≦40 mm), resulting in behaviors that differ significantly from those of the in situ soil and causing the obtained conditioning scheme to perform poorly during actual tunnel construction. This study applied a laboratory-obtained conditioning scheme during EPBS tunneling in the boulder- and pebble-rich soil strata between Yushuzhuang and Wanpingcheng on Beijing Metro Line 16 to determine the optimal soil conditioning scheme using the upper soil chamber pressure, cutterhead torque, tunneling speed, and total thrust tunneling parameters as evaluation indices. The optimized soil conditioning scheme provided a better soil conditioning effect than the laboratory-obtained scheme and confirmed that the considered parameters reflected the soil conditioning effects. Finally, the correlations between these three soil conditioning factors and the four tunneling parameters were analyzed using a full factorial experimental design to obtain contour plots of their quantitative relationships for use in similar tunneling projects.

1. Introduction

The soils of boulder- and pebble-rich strata primarily comprise large-sized pebbles and boulders and lack fine particles such as clay and silt. This composition can lead to difficulties in building up pressure in the soil chamber, excess muck discharge, and surface settlement when constructing EPBS tunnels. Furthermore, the strongly abrasive properties of large-size pebbles and boulders cause severe cutter wear, necessitating frequent opening of the soil chamber for repairs, which leads to delays in the project schedule and increases cost [1]. Numerous engineering examples have shown that reasonable soil conditioning is an effective means of addressing the issues associated with EPBS tunneling in such boulder-rich strata. Methods for soil conditioning include injecting bentonite, foam [2,3], or other agents to effectively improve the workability [4,5] and flow plasticity of the soil, ensure stability of tunnel excavation, and promote the reduction of cutter wear [6]. Therefore, ongoing research on soil conditioning technology for boulder-rich pebble strata is necessary to realize safe and cost-effective EPBS tunnel construction.

Laboratory tests are the most commonly used methods for studying the effects of soil conditioning. Peila et al. [7] developed a prototype laboratory device to simulate the process of muck discharge from the soil chamber to the screw conveyor, Peila [8] subsequently applied this device to investigate the foam-conditioned sand layer. Xu et al. [9] investigated the effect of soil conditioners on clogging in EPB through laboratory experiments. Wang et al. [10] analysed the effects of organic and inorganic dispersants on clay plasticity and slaking characteristics through Atterberg limit and slaking tests. Huang et al. [11] used an improved push-pull dynamometer to investigate the influence of gravimetric water content, Foam injection ratio (FIR), fines content, and clay mineralogy on the compression-adhesion behavior of foam-conditioned soil. Lee et al. [12,13] established a laboratory-scale excavation test apparatus to investigate the excavation performance of EPB in foam-conditioned weathered granite soil. Wan et al. [14] proposed a composite soil conditioning scheme for tunneling in highly viscous clay and weathered mudstone strata using anti-clay and foam agents. This approach was effectively applied in actual construction. Sebastiani et al. [15] adopted Hobart mixing and the plate pull-out test to evaluate the influnce of grain size distribution and mineralogical content for clogging tendency of natural clay. Ling et al. [16] developed a novel calculation model through permeability tests to estimate the permeability coefficient of soils conditioned with foam and bentonite slurry. Li et al. [17] experimentally investigated the interactions between foam agents and polymers to propose a soil conditioning scheme for water-rich sand layers. Wan et al. [18] analyzed the contact relationships between different particles in foam-conditioned soil via microscopic experiments, and simulated its mechanical behavior during transportation in a screw conveyor using the DEM. Yang et al. [19,20] investigated the injection ratio for water-rich sand layer conditioning and evaluated the characteristics of the conditioned soil using indoor experiments. Finally, Carigi et al. [21] proposed a new method for assessing the homogeneity of conditioned cobble flow and introduced two indices to evaluate the suitability of the applied conditioning measures.

Scholars globally have conducted extensive research on soil conditioning for EPBS tunneling in pebble- and gravel-rich strata from different perspectives, achieving suitable application results. Barzegari et al. [22] proposed the use of a high foam expansion rate (FER) foam to condition the soil of a boulder-rich stratum along Line 1 of the Tabriz Metro that effectively reduced the cutterhead torque and prevented surface collapse. Zumsteg and Langmaack [23] proposed the use of foam, polymer, and bentonite for the soil conditioning of Swiss glacial deposits that were rich in gravel. This scheme was proven successful when applied in an actual project. Zhang et al. [24], Jiang et al. [25], and Zhang et al. [26] studied the soil conditioning of sand and pebble-rich strata in Lanzhou, Beijing, and Chengdu, China, respectively, and proposed optimal injection ratios for bentonite, foam, and polymer accordingly. Zhen et al. [27] proposed the optimal ratio of foam for EPB tunneling in sandy and cobbly soil through laboratory and field experiments. Wang et al. [28] discussed the influence of the auxiliary air pressure balance mode and earth pressure balance mode of EPB in water-rich gravelly sand strata on the feasibility and soil conditioning effect. Huang et al. [29], Wang et al. [30] investigated the effect of gradation on undrained compressibility and permeability of foam-conditioned coarse-grained soils. Researchers have also studied the soil conditioning of larger pebble- and boulder-rich strata. Wei et al. [31] studied soil conditioning in sand-cobble strata rich in large-grained pebbles and concluded that conditioning schemes for such soils must be adapted to local conditions through field tests. Notably, Yao et al. [32] proposed a hybrid experimental–theoretical method for conditioning strata containing large pebbles to correct the results obtained from soil conditioning tests on samples in which particles larger than 30 mm were removed.

While scholars have achieved valuable results for the conditioning of soil strata with large pebbles and boulders, most of the experimental research conducted to date replaced these larger elements with an equal mass of smaller pebbles (generally ≦40 mm in diameter). This leads to a significant difference between the behaviors of the experimental and in-situ soils and considerably reduces the applicability of the laboratory test results. For instance, in the Jun-Dong section of Beijing Line 9, due to the direct adoption of indoor experimental results to set the soil conditioning plan, the EPBs encountered the cutterhead jamming and severe wear of the cutters during tunneling in the boulder stratum (Figure 1). Therefore, laboratory testing methods for evaluating the conditioning of soil strata containing large pebbles and boulders must be further improved.

This study addressed the limited applicability of soil conditioning schemes obtained by laboratory testing of strata rich in pebbles and boulders by first applying a laboratory-obtained soil conditioning scheme during the process of EPBS tunnel construction in the section of Beijing Metro Line 16 from Yushuzhuang to Wanpingcheng. The optimal soil conditioning scheme was subsequently determined by adjusting the slurry, foam, and mobile agitator soil conditioning factors to optimize four evaluation indexes established based on the upper soil chamber pressure, cutterhead torque, tunneling speed, and total thrust tunneling parameters. Finally, the correlations between the three considered soil conditioning factors and four tunneling parameters were analyzed using a full factorial experimental design.

2. Tunneling Conditions

2.1. Overview

The Yushuzhuang–Wanpingcheng section of Beijing Metro Line 16 is located in the southwest of Beijing’s Fengtai district. It comprises two parallel EPBS tunnels running 2800 m from Yushuzhuang station to Wanpingcheng station, as shown in Figure 2. The applied tunnel shield consists of C50 reinforced concrete ring segments with an outer diameter of 6400 mm, inner diameter of 5800 mm, and length of 1200 mm. This study primarily focused on the soil conditioning scheme for the inbound (right) tunnel.

2.2. Geology

The right tunnel between Yushuzhuang and Wanpingcheng Stations of Beijing Metro Line 16 has a burial depth of approximately 9–20 m. As shown in Figure 3, it primarily passes through the boulder-and pebble-rich layers ③ and ④ and locally passes through a clay layer ⑦. The groundwater level is located below the tunnel base plate. According to the results obtained from manually excavated exploration wells by Yang et al. [1], the mass fraction of grains larger than 200 mm in pebble layer ③ ranges from 30% to 42%, with the largest grain size being 420 mm. The mass fraction of huge grains larger than 200 mm in pebble layer ④ is approximately 34% to 46%, and the largest grain size is 670 mm. According to the engineering classification standard for soils (2007) huge grains with a grain size larger than 200 mm are typically defined as boulders. In this soil layer, the content of boulders is more than 30% in all cases. Thus, this is a typical stratum rich in large-grained pebbles and boulders. The uniaxial compressive strengths of the boulders ranged from 177 to 448 MPa with a Cerchar abrasivity index of 2.08–3.49, indicating highly abrasive strata that pose considerable challenges to EPBS tunnel construction.

2.3. EPBS Tunneling Equipment

The EPBS equipment used to construct the subject tunnel comprised a composite cutterhead with four main beams and four panels, as shown in Figure 4. It provided a cutter excavation diameter of 6680 mm, an opening rate of 46%, a rated torque of 7850 kN·m, and a jam breakout torque of 9500 kN·m. The range of grain sizes allowed to pass through this cutterhead was 700 mm to 1200 mm. The four main beams of the cutterhead were equipped with cutters and scrapers, and each panel was equipped with welded heavy-duty rippers inlaid with tungsten carbide blocks to improve impact and wear resistance. The cutterhead arrangement was categorized into three layers: the first layer comprised 175-mm high disc cutters installed on the main beams, the second layer consisted of 155-mm high heavy-duty rippers welded to the panels and ends of the main beams, and the third layer comprised scrapers with 120-mm long cutters installed along the edges of the main beams. The cutterhead was equipped with six soil-conditioning agent injection holes that could be used to inject foam or bentonite slurry. In addition, two slurry-only injection holes in the soil chamber partition permitted the injection of high-viscosity bentonite slurry into the soil chamber.

3. Laboratory Tests to Determine Soil Conditioning Scheme

3.1. Bentonite Slurry Concentration

As bentonite is primarily composed of clay minerals, its addition to the surrounding soil can supplement the content of fine particles within to improve mobility and cohesion by wrapping larger particles, such as those found in pebble- and boulder-rich strata. The Marsh funnel (106 type) test was used to evaluate the relationship between the slurry viscosity index and expansion time based on bentonite slurry concentration, as shown in Figure 5. The viscosity of the bentonite slurry gradually increased as its bentonite concentration increased from 6% to 9% and sharply increased above 9%. Bentonite should provide excellent workability as well as viscosity at higher concentrations. A bentonite slurry concentration of 9% provided an expansion time of 12 h and a marsh funnel viscosity of 90 s.

3.2. Foam Properties

Foam is a two-phase gas-liquid material produced by mixing a small quantity of foaming agent solution with a large quantity of air. The foaming agent in this study contained a surfactant [33] to reduce the friction between particles and improve the fluidity of the muck. However, the structure of a foam is unstable and can easily fail over time. Therefore, half-life times (H-Ts) of different foam agent concentrations were determined through laboratory testing, as shown in Figure 6. A 5% foam concentration was determined to provide the best performance with an FER of 37 and an H-T of 35 min.

3.3. Muck Slump Test

The slump test is a simple and intuitive method for evaluating the workability of the muck produced during EPBS tunneling. Notably, laboratory slump test results will be inconsistent with the actual effect of in situ soil conditioning in strata containing large pebbles and boulders because the small volume of the slump bucket can only accommodate cobbles with a maximum particle size of 40 mm. However, this test remains useful for an initial determination of the conditioning agent dosage range prior to the beginning of tunneling. In this study, an original pebble layer ③ soil sample excavated at Yushuzhuang station was used for slump testing, with results presented in

Table 1. Slump testing of soil conditioning schemes.

Test Number	Soil Conditioning Agent Volumetric Injection Ratio		Slump (cm)	Flow Plasticity
	SIR (%)	FIR (%)
1	6		15.6	Loose and poor
2	8		13.4	Poor
3	10	/	11.7	Good
4	12		8.8	Medium
5	14		8	Muck water precipitation
6		6	11.8	Poor
7		16	13.5	Poor
8	10	26	17.5	Good
9		36	18.6	Good: muck began to precipitate water
10		46	19.8	Muck water precipitation

and Figure 7. In these tests, pebbles and boulders larger than 40 mm were removed from the soil and replaced with the same mass of pebbles sized 30 to 40 mm. First, an optimum bentonite slurry injection ratio (SIR) with no foam was determined by slump testing, then different FIRs were added to determine the optimum combined dosage of bentonite slurry and foam by further slump testing.

3.4. Results of Laboratory Tests

The optimal concentration of each soil conditioning agent was determined individually based on the results of the property tests as presented in Section 3.1 and Section 3.2. The slump test results for different soil conditioning agent volumetric ratios were subsequently used to determine the optimal combination of agents as presented in Section 3.3. The optimal soil conditioning scheme for the sample evaluated in the laboratory tests comprised a 10% SIR with a 9% bentonite concentration and a 26% FIR with a 5% foam concentration.

4. Field Testing

4.1. Test Sections and Soil Conditioning Schemes

The field test initially applied the laboratory-obtained design soil conditioning scheme, then adjusted this scheme based on the observed effects during EPBS tunneling to ultimately obtain the optimal soil conditioning scheme. The optimal soil conditioning scheme was not obtained until the tunnel had progressed through 500 rings of length (600 m) owing to the large grain size and extremely uneven grain distribution in the strata traversed in this project. The entire field test can be categorized into four stages: the first three stages comprised 100 rings each (120 m) and the fourth stage comprised 200 rings (240 m). After the soil conditioning scheme was optimized over the first three stages, it exhibited a suitable field application effect in the fourth stage. The soil conditioning schemes applied in the four stages are detailed in

Table 2. Soil conditioning schemes applied in the field test.

Test Stage	Length (Rings)	Number of Mobile Agitators	Bentonite Injection Volume	Foam Injection Volume	Upper Soil Chamber Pressure	Average Cutterhead Torque	Average Tunneling Speed	Average Total Thrust
1	100	4	4.6 m3/rings	12 m3/rings	0.22 bar	5800 kN·m	24 mm/min	15,000 kN
2	100	4	9 m3/rings	20 m3/rings	0.5 bar	3998 kN·m	49 mm/min	13,915 kN
3	100	4	6 m3/rings	16 m3/rings	0.7 bar	4995 kN·m	59 mm/min	16,282 kN
4	200	6	6 m3/rings	16 m3/rings	0.8 bar	3154 kN·m	71 mm/min	9398 kN

, along with the corresponding upper soil chamber pressure, cutterhead torque, tunneling speed, and total shield thrust parameters used to characterize the soil conditioning effect. The curves for these parameters are shown in Figure 8a–d. Note that the schemes in the table are reported in terms of volume of agent per ring determined using the diameter of the shield excavation (for a soil volume of 46 m³/rings) and the desired SIR and FIR.

4.2. Analysis of Soil Conditioning Effects

4.2.1. Stage 1 (Rings 1–100)

Stage 1 comprised the field application of the laboratory-obtained design soil conditioning scheme, presented in Section 3.4: a 10% SIR with a 9% bentonite concentration and a 26% FIR with a 5% foam concentration, corresponding to a bentonite slurry injection volume of 4.6 m³/rings and a foam injection volume of 12 m³/rings. As shown in Figure 8a–d, during the tunneling of rings 1–100, the upper soil chamber pressure fluctuated considerably within 0.05–0.39 bar with an average of 0.22 bar, reflecting difficulties establishing a stable soil chamber. The cutterhead torque was large at 3400–8200 kN·m with an average of 5836 kN·m, frequently exceeding the rated torque and leading to many emergency stops. The tunneling speed was slow at 5–40 mm/min with an average of 24 mm/min. Finally, the total thrust during shield tunneling ranged from 13,000–18,000 kN with an average of 15,082 kN.

The muck discharge from the screw conveyor during Stage 1 exhibited low water content and was sometimes excessive discharge. Notably, three collapses occurred on the ground surface during this stage, as shown in Figure 9.

4.2.2. Stage 2 (Rings 101–200)

Considering the difficulties encountered during Stage 1, which included fluctuations in the soil chamber pressure, a large torque on the cutterhead, and a low muck water content, the injection volumes of bentonite slurry and foam were increased from 4.6 to 9 m³/rings and from 12 to 20 m³/rings, respectively, in Stage 2.

Comparing Stages 1 and 2 in Figure 8a–d, during the excavation of rings 101–200, the upper soil chamber pressure on the shield increased to fluctuate within a smaller range of 0.4–0.6 bar with an average of 0.52 bar, which was generally still low. The cutter torque was 2200–5500 kN·m with an average of 3998 kN·m, representing a 31.5% reduction compared to that in Stage 1. The tunneling speed was 40–60 mm/min with an average of 49 mm/min, which was twice as high as that in Stage 1. The total thrust of the shield was 12,000–17,000 kN with an average of 13,915 kN, exhibiting a slight decrease from that in Stage 1.

Stage 2 exhibited a high water content in the muck discharge from the screw conveyor and the discharge of muck still exceeded the expected discharge volume. One ground surface collapse occurred.

4.2.3. Stage 3 (Rings 201–300)

Although the modified soil conditioning scheme applied in Stage 2 significantly improved the tunneling parameters over those obtained in Stage 1, problems such as the higher water content of the muck, low soil chamber pressure, and surface collapse remained. Therefore, the soil conditioning scheme was further optimized in Stage 3 by reducing the bentonite slurry injection volume from 9 to 6 m³/rings and the foam injection volume from 20 to 16 m³/rings.

Comparing Stages 2 and 3 in Figure 8a–d, during the excavation of rings 201–300, the upper soil chamber pressure on the shield increased to 0.6–0.8 bar with an average of 0.71 bar. Furthermore, the cutterhead torque increased to 3000–7500 kN·m with an average of 4995 kN·m owing to the decrease in the volume of foam injected. The tunneling speed increased to 65–80 mm/min with an average of 59 mm/min, and the total thrust increased to 15,000–19,000 kN with an average of 16,282 kN.

The muck discharge from the screw conveyor was unstable and frequently exhibited either insufficient or excessive water content, but the phenomenon of excess muck discharge disappeared. No ground surface collapses occurred.

4.2.4. Stage 4 (Rings 301–500)

The soil conditioning effect was improved in Stage 3, but the cutterhead torque and muck discharge characteristics still required improvement. As these issues were determined to result from the uneven mixing of the muck by the agitators, the shield mixing system configuration was optimized when the system reached the inspection shaft located at ring 300. One pair of mobile agitators was added to the outermost panel on the back of the cutterhead, the original mobile agitators were lengthened, and one pair of fixed agitators was eliminated from the original soil chamber partition, as shown in Figure 10. The modified soil conditioning scheme applied in Stage 3 was also used in Stage 4.

Comparing Stages 3 and 4 in Figure 8a–d, during the excavation of rings 301–500, the upper soil chamber pressure on the shield increased to 0.7–0.9 bar with an average of 0.79 bar. Furthermore, the cutterhead torque was dramatically reduced to 1500–6000 kN·m with an average of 3154 kN·m. The tunneling speed range was nearly unchanged at 65–85 mm/min but the average increased to 71 mm/min. Finally, the total thrust dramatically decreased to 8000–12,000 kN with an average of 9398 kN.

The muck discharge from the screw conveyor was excellent in terms of water content and quantity, and no surface collapses occurred. Although the same soil conditioning scheme was used in Stages 3 and 4, the tunneling parameters and muck discharge conditions indicated that the muck was more adequately mixed with the conditioners when the modified mixing system was applied in Stage 4. Indeed, the upper soil chamber pressure increased by 11.3%, the cutterhead torque decreased by 36.9%, and the total thrust decreased by 42.3% from Stage 3 to 4, indicating that improved muck mixing can effectively enhance the soil conditioning effect.

4.3. Correlation Between Soil Conditioning Factors and Tunneling Parameters

The full factorial design (FFD) is a multifactorial experimental design method that considers all possible combinations of factors to comprehensively assess the effects of each on the results. This method can be used to study the effects of changes in multiple experimental factors on the experimental indicators and thereby identify the most significant factor. An FFD was applied in this study using the soil conditioning scheme factors (injected bentonite slurry volume, injected foam volume, and number of mobile agitators) and tunneling parameters (upper soil chamber pressure, cutterhead torque, tunneling speed, and total thrust) to assess the extent to which each of the former affected the soil conditions represented by each of the latter. Two levels (±1) were examined for each parameter, where the high level (+1) was taken as 1.2 times the mean value of the experimental design factor and the low level (−1) was taken as 0.8 times the mean value of the experimental design factor. The test design and results are listed in

Table 3. Full factorial test scheme and results.

Operational Sequence	Bentonite Slurry Volume (m3)	Foam Volume (m3)	Number of Mobile Agitators *	Upper Soil Chamber Pressure (bar)	Cutterhead Torque (kN·m)	Tunneling Speed (mm/min)	Total Thrust (kN)
1	6	12	4	0.71	4955	69	16,282
2	4.6	20	6	0.52	3998	49	13,915
3	5.3	16	5	0.74	5025	62	18,762
4	4.6	12	6	0.22	5836	24	15,082
5	6	12	4	0.74	4722	59	17,620
6	6	20	6	0.79	3154	71	9398
7	4.6	20	4	0.55	4027	47	14,015
8	5.3	16	5	0.69	4725	67	15,994
9	6	20	4	0.72	4625	68	15,823
10	4.6	12	4	0.21	6020	22	15,432
11	4.6	20	4	0.53	3888	49	13,765
12	6	12	6	0.74	3244	70	9574
13	5.3	16	5	0.72	4877	69	15,342
14	6	12	6	0.77	3241	72	9798
15	5.3	16	5	0.71	4799	69	15,300
16	6	20	4	0.72	4625	68	15,823
17	4.6	12	4	0.21	6020	22	15,432
18	4.6	12	6	0.22	5836	24	15,082
19	6	20	6	0.79	3154	71	9398
20	5.3	16	5	0.72	4800	69	15,322
21	4.6	20	6	0.52	3998	49	13,915

* Note: Because the original agitators were extended, the extended mobile agitators were counted as 1.5 each instead of 1.

.

The FFD model was solved to obtain the normal distribution between each soil conditioning factor and tunneling parameter along with the standardized effect Pareto plots, as shown in Figure 11. In Figure 11a,c, the bentonite slurry volume exerted an extremely significant effect on the upper soil chamber pressure and tunneling speed, respectively, with obvious positive correlations. Figure 11b indicates that all three soil conditioning factors—the bentonite slurry volume, foam volume, and number of mobile agitators—had a significant effect on the cutterhead torque with negative correlations. The foam volume had the greatest effect. As illustrated in Figure 11d, the number of mobile agitators and bentonite slurry volume exhibited significant negative correlations with the total thrust. Contour plots of the relationships between the soil conditioning factors and shield tunneling parameters are displayed in Figure 12 to provide a reference for the determination and adjustment of soil conditioning schemes in the field.

5. Discussion

The optimization of the soil conditioning scheme based on the resulting tunneling parameters observed during the field test provided a better soil conditioning effect than the conventional laboratory tests and confirmed that the upper soil chamber pressure, cutterhead torque, tunneling speed, and total thrust parameters reflect the effects of soil conditioning.

Notably, soil chamber pressure is a key parameter for evaluating the soil conditioning effect. Establishing a stable soil chamber pressure can be difficult owing to the extremely uneven particle size distribution and poor workability of the muck from strata containing large pebbles and boulders. As unstable soil chamber pressure can lead to destabilization of the tunnel face, ground collapse, and other accidents, the stability of the soil chamber pressure is a critical indicator of the effectiveness of soil conditioning to ensure tunnel face stability. The cutterhead torque is another critical parameter, as strata containing large pebbles and boulders can be extremely abrasive, easily causing cutterhead jams and severe cutter wear. Properly designed soil conditioning can effectively reduce the frictional resistance of a stratum while improving the workability and mobility of the muck, thereby reducing the cutterhead torque. Finally, the tunneling speed and total thrust provide intuitive reflections of the advantages and disadvantages of the applied soil conditioning scheme. When the effect of the soil conditioning scheme is poor, the shield tunneling speed decreases and the total thrust increases. Conversely, when the effect of the soil conditioning scheme is good, the tunneling speed increases and the total thrust decreases.

6. Conclusions

The Yushuzhuang–Wanpingcheng section of Line 16 of the Beijing Metro was selected as the subject of this study to conduct field tests of soil conditioning schemes when constructing EPBS tunnels in strata rich in large boulders and pebbles. The primary conclusions of this study are as follows:

1.

The soil chamber pressure, cutterhead torque, tunneling speed, and total thrust parameters can provide feedback on the effect of soil conditioning schemes in real time, offering a more practical and effective approach than laboratory testing.

2.

The field test conducted in this study provided a well-adapted and practically tested soil conditioning scheme according to the geological conditions of the strata at the project site. In this scheme, a 34.6% FIR with 9% foam was injected into the soil in front of the cutterhead, and a 13% SIR with 5% bentonite was injected into the soil chamber with excellent results in terms of the four evaluated tunneling parameters. This result provides a technical reference for the development of soil conditioning schemes when constructing EPBS tunnels in similar strata.

3.

Given the same soil conditioning agent injection ratios, optimizing the shield mixing system increased the pressure in the upper soil chamber by 11.3%, decreased the torque on the cutterhead by 36.9%, and decreased the total thrust by 42.3%. Thus, improving the mixing capacity of the soil chamber can effectively improve the effects of soil conditioning.

4.

The correlations between the three considered soil conditioning factors (bentonite slurry injection, foam injection, and number of mobile agitators) and the four tunneling parameters (upper soil chamber pressure, cutterhead torque, tunneling speed, and total thrust) were analyzed using an FFD to derive their quantitative relationships. These relationships, presented in the form of contours, provide a reference for determining the dosages of soil conditioning agents during EPBS tunneling in similar strata.