Performance and Microstructure of Grouting Materials Made from Shield Muck

In response to the environmental pollution caused by transportation and accumulation of large-scale shield muck, the on-site reutilization of shield muck is an effective approach. This study explored the feasibility of silty clay muck to prepare muck grout. Through orthogonal experiments, the effects of cement, fly ash, shield muck, admixture, and the water–solid ratio on the fresh properties and mechanical properties of muck grout were studied. The performance prediction model was established Additionally, the intrinsic relationships between the compressive strength and microstructure of shield muck grouting materials were explored through multi-technology microstructural characterization. The results indicate that the content of muck and the water–solid ratio have a greater significant influence on the bleeding ratio, flowability, setting time, and volume shrinkage rate of muck grout compared to other factors. Cement has a greater significant influence on the compressive strength of muck grout than other factors. An optimal mix proportion (12% for cement, 18% for fly ash, 50% for muck, 0.465 for water–solid ratio, 19.5% for river sand, and 0.5% for bentonite) can produce grouting materials that meet performance requirements. The filling effect of cementitious substances and the particle agglomeration effect reduce the internal pores of grouting materials, improving their internal structure and significantly enhancing their compressive strength. Utilizing shield muck as a raw material for shield synchronous grouting is feasible.


Introduction
With the rapid development of underground space, shield tunneling has been widely used in construction projects such as rail transportation, municipal highways, and urban comprehensive pipe corridors due to its advantages of high efficiency, safety, and mechanization [1,2].However, the shield construction process generates a massive amount of muck.According to incomplete statistics, the annual total volume of shield muck generated by shield tunnels in China has exceeded 225 million cubic meters, with disposal costs estimated to reach CNY 58.2 billion [3].Shield muck has characteristics such as small particle size, high water content, and poor stability, making it difficult to be completely recycled using traditional construction waste methods.Currently, shield muck is still treated by landfill and stacking [4,5], which not only occupies land resources and increases costs but also causes severe environmental pollution such as dust, noise, and spillage during transportation.Although scholars have used shield muck to produce sintered bricks [6], high-belite cementitious materials [7], recycled aggregates [8], and subgrade materials [9], the actual utilization rate of shield muck in China is less than 1% [10].The disposal and utilization of large volumes of shield muck have become a major challenge hindering the safe and efficient construction of shield tunnels.
In recent years, the use of shield muck to prepare grouting materials has become a research hotspot.This approach can effectively prevent environmental pollution and disasters such as landslides caused by shield muck accumulation, while significantly reducing carbon emissions during the extraction and processing of grouting raw materials [11].
Vinai et al. explored the changes in slump and strength of shield muck under different water content and foam content through laboratory tests [12].Zhou and Zhang successfully applied shield muck treated by the slurry system in backfill grouting [13,14].Wang et al. found that the slurry prepared from sandy muck with appropriate material proportions can meet the performance requirements of synchronous grouting and validated the effectiveness of muck grout through monitored values such as surface settlement and segment uplift [15][16][17].Although there have been reports on the application of shield muck in synchronous grouting, most focus on sandy soil or silty clay in slurry shields, with less research on the application of silty clay in earth pressure balance shields (the muck in slurry shields differs from that in earth pressure shields due to differences in properties).Given the different particle compositions of silty clay and sandy soil, the suitability of grouting materials must differ [18,19].Some studies have shown that partially replacing river sand with silty clay reduces the flowability of backfill grouting and extends setting time [20,21], while high water content can lead to increased bleeding and reduced strength [22].Therefore, further research is needed to prepare grouting materials from silty clay muck that meet the performance requirements of synchronous grouting.
This paper takes the silty clay muck from an excavation tunneling section of the Yinhan Jiwei Project as an example to explore the feasibility of using silty clay muck as a synchronous grouting raw material.Through orthogonal experiments, the fresh and mechanical properties of shield muck grouting materials (SMGM) with different material proportions were tested.The impact of material composition on their performance was analyzed, and a performance prediction model was established.Based on the performance requirements, suitable mix proportions for grouting materials were recommended.Additionally, the solidification mechanism of the SMGM was described at the microscopic level using various techniques such as XRD, SEM, and CT.This study provides a reference for the application of shield muck in synchronous grouting.It is of great significance for reducing costs, improving construction efficiency, and minimizing the environmental impact of synchronous grouting materials.

Experimental Scheme 2.1. Materials
The soil samples were obtained from a shield construction site in Xianyang City, Shanxi province of China.It was bagged and transported to the laboratory for natural air drying and was mechanically sieved through a 2 mm sieve mesh.According to GB50123-2019 [23], the relevant physical properties were tested.Its average values of dry density, plastic limit, liquid limit, and plasticity index are 1.55 g/cm 3 , 18.9%, 33.8%, and 14.9, respectively.The soil layer in this area is loess, primarily composed of silt and clay, characterized by high porosity and permeability.
The cementitious materials were P.O 32.5 ordinary Portland cement and fly ash produced in Shijiazhuang City, Hebei province of China.The bentonite was calcium-based bentonite.According to the data provided by the manufacturer, the chemical composition of cement and fly ash is shown in Table 1, and the detailed parameters of bentonite are listed in Table 2.The fine aggregate was natural river sand, with an apparent density of 2600 kg/m 3 and a fineness modulus of 2.79.The particle size distributions of loess, river sand, and bentonite are shown in Figure 1.Loess and bentonite have a high clay content.The mineral composition was determined using XRD testing, and the results are presented in Figure 2. The main minerals of river sand, bentonite, and loess are quartz, which is stable in nature and generally does not react with other substances [24].Besides quartz, loess and river sand contain stable minerals such as feldspar, calcite, and muscovite.Therefore, from the perspective of mineral composition, the muck can be used as a raw material for backfill grouting.The admixture was a polycarboxylate superplasticizer with a water reduction rate of 25% and sodium tetraborate, which has a retarding effect.Based on preliminary tests, the composition of the admixture used in this study was superplasticizer:tetraborate = 1:1.a water reduction rate of 25% and sodium tetraborate, which has a retarding effec on preliminary tests, the composition of the admixture used in this study was su ticizer:tetraborate = 1:1.

Experimental Design
Based on relevant studies and preliminary tests [25][26][27], an orthogonal exp was conducted to design the mix proportions for SMGM [28].The orthogonal exp factors included the content of cement, fly ash, shield muck, admixture, and the solid ratio, with each factor set at four levels as shown in Table 3.The amount of ad was expressed as a percentage of the cementitious materials.The L16(4 5 ) orthog sign was used for the experiment.Relevant tests were carried out on 16 groups of and the test results are shown in Table 4. a water reduction rate of 25% and sodium tetraborate, which has a retarding effect.Based on preliminary tests, the composition of the admixture used in this study was superplas ticizer:tetraborate = 1:1.

Experimental Design
Based on relevant studies and preliminary tests [25][26][27], an orthogonal experimen was conducted to design the mix proportions for SMGM [28].The orthogonal experimen factors included the content of cement, fly ash, shield muck, admixture, and the watersolid ratio, with each factor set at four levels as shown in Table 3.The amount of admixture was expressed as a percentage of the cementitious materials.The L16(4 5 ) orthogonal de sign was used for the experiment.Relevant tests were carried out on 16 groups of SMGM and the test results are shown in Table 4.

Experimental Design
Based on relevant studies and preliminary tests [25][26][27], an orthogonal experiment was conducted to design the mix proportions for SMGM [28].The orthogonal experiment factors included the content of cement, fly ash, shield muck, admixture, and the water-solid ratio, with each factor set at four levels as shown in Table 3.The amount of admixture was expressed as a percentage of the cementitious materials.The L16(4 5 ) orthogonal design was used for the experiment.Relevant tests were carried out on 16 groups of SMGM, and the test results are shown in Table 4.

Grout Preparation
The preparation of the SMGM was conducted as follows.Initially, the necessary dry materials (cement, fly ash, muck, sand, and bentonite) were added to the mixing pot and stirred for 3 min.Subsequently, water mixed with an admixture was combined with the dry ingredients.Stirring was continued for an additional 3 min to produce the SMGM.It was important to note that the moisture content of the shield muck and river sand was considered when calculating the water-solid ratio.

Test Methods
This paper tests the performance of the SMGM from three aspects: fresh properties, mechanical properties, and microstructure properties, as shown in Figure 3.The fresh properties of the SMGM include bleeding ratio, setting time, and flowability.The bleeding ratio of the SMGM was tested according to T/CECS563-2018 [29].The

Fresh Property Tests
The fresh properties of the SMGM include bleeding ratio, setting time, and flowability.The bleeding ratio of the SMGM was tested according to T/CECS563-2018 [29].The fresh slurry was poured into a graduated cylinder of 250 mL and sealed with plastic film.After allowing the slurry to stand for 1 min, the initial volume of the slurry was recorded.The bleeding ratio was calculated as the ratio of the volume of water separated from the slurry after 3 h to the initial volume of the slurry.The setting time of the SMGM was tested according to the JGJ/T70-2009 [30].The setting time was determined by the penetration resistance method using a ZKS-100 setting time tester.The container holding the slurry was placed on the pressure gauge dial.The pressure gauge needle was set to zero by adjusting the nut.The penetration needle was then brought into contact with the surface of the slurry and uniformly pressed into the slurry to a depth of 25 mm within 10 s.The gauge reading was recorded each time the needle penetrated.The setting time of the slurry was determined when the penetration resistance reached 15 N.The flowability of the SMGM was tested according to the GB/T 50448-2008 [31].The test apparatus was a truncated cone mold with a 70 mm top diameter, 100 mm bottom diameter, and 60 mm height.The mold was placed on a 50 × 50 cm glass plate.First, fresh slurry was poured into the mold.Then, the mold was lifted vertically to allow the slurry to flow freely for 30 s.The average of the maximum diffusion diameter and its perpendicular diameter is taken as the flowability of the slurry.In addition, according to the actual engineering application, it takes 0-3 h from the completion of fresh grout to be injected into the shield tail gap.Therefore, the flowability of the SMGM was tested every 1 h after the fresh grout was complete.It was stirred again for 3 min before each test.

Mechanical Property Tests
The mechanical properties of the SMGM include volume shrinkage rate and compressive strength.The volume shrinkage rate was calculated as the ratio of the difference between the actual volume of the sample and the standard volume of the mold to the standard volume of the mold [32].The compressive strength of the SMGM was tested according to JGJ/T70-2009 [30] by using the YHS-229WJ electronic universal testing machine.The sample size was 70.7 mm × 70.7 mm × 70.7 mm.The sample was placed at the center of the pressure platform, ensuring the upper platen was in contact with the sample surface.The loading rate was adjusted to 1 mm/min, and loading was continued until the sample failed.The compressive strength of the sample was then recorded.The compressive strength of the SMGM was tested after 1 d, 3 d, 7 d, and 28 d of curing.Three parallel samples were used, and the average value was taken as the compressive strength.

Microstructure Property Tests
The microstructure of the SMGM hardened for 28 d was characterized by using XRD, SEM, and X-CT.The mineral composition of the SMGM was examined by using an SmartLab(9)kW XG X-ray diffraction instrument (Rigaku Corporation, Tokyo, Japan).Samples were placed in an oven at 105 • C to dry, then crushed to a size of less than 75 µm and placed in the instrument.The relevant parameters were adjusted (the diffraction angle 2θ scan range was set to 5 • to 90 • , and the scan speed was set to 5 • /min) before conducting the test.The micromorphology of the SMGM was investigated by using a Phenom pure scanning electron microscope (Phenom-World, Eindhoven, The Netherlands).Samples were dried in an oven at 105 • C.They were adhered to the observation platform using conductive adhesive, and surface dust was blown off using an ear syringe before being placed in the instrument for testing.The pore characteristics of the SMGM were examined by a nano Voxel 2740 high-resolution X-ray digital core computed tomography (Sanying Precision Instruments Co., Ltd., Tianjiin, China).The sample size was ϕ50 mm × 50 mm.Samples were placed on the CT bearing platform, relevant parameters were adjusted, and scanning was started.After scanning, visualization software was used for threedimensional reconstruction and analysis of the pore structure.

Performance Requirements
The purpose of synchronous grouting in shield tunneling is to prevent segment uplift, reduce surface settlement, and enhance the overall stability of the tunnel [33,34].Therefore, the grouting slurry should have good conveyance performance, filling performance, and suitable strength.This study focuses on the synchronous grouting in loess strata.Through laboratory tests, the compressive strength of the undisturbed soil was found to be 0.11 MPa.By consulting specifications, the literature, and on-site construction conditions, the performance requirements for the grouting material used in this project are as follows: 3 h flowability greater than 160 mm, setting time between 10 and 24 h, bleeding ratio less than 5%, volume shrinkage rate less than 5%, and compressive strengths of not less than 0.15 MPa at 3 d and 1 MPa at 28 d.

Influence of Factors on Bleeding Ratio
Figure 4 shows the variation in the bleeding ratio of the SMGM.The bleeding ratio of the SMGM could be adjusted over a wide range (1.2~11.79%)by changing the mix ratio of grout, with NO. 7 having the lowest bleeding ratio (1.2%) and NO. 6 having the highest (11.79%).The bleeding ratios of NO. 2~NO.4, NO.7~NO.10, and NO.12~NO.15 were below 5%.The influence of five factors on the bleeding ratio is shown in Figure 5.The range analysis result presented that the significance of each factor on the bleeding ratio was ranked in the order of R C > R E > R A > R B > R D .It showed that muck content had a great influence on the bleeding ratio of the SMGM (R C = 7.51).The bleeding ratio of the SMGM decreased with the increase in the content of muck and fly ash.This is because clay minerals such as montmorillonite in the muck have a strong ability to adsorb water molecules [35].As the muck increased, more water molecules were adsorbed by soil particles, the amount of free water molecules was reduced and thus the bleeding ratio was lowered.On the other hand, the bleeding ratio of the SMGM increased with the increase in the water-solid ratio.This is because as the water-solid ratio increased, more free water was not used for the hydration reaction, causing an increase in the bleeding ratio of the SMGM [36].When the cement content was less than 8%, the bleeding ratio of the SMGM increased with the increase in cement content.However, when the cement content was greater than 8%, the bleeding ratio of the SMGM decreased with the increase in cement content.The admixture had the least impact on the bleeding ratio of the SMGM.

Influence of Factors on Setting Time
Figure 6 shows the variation in the setting time of the SMGM.The setting time of the SMGM could be adjusted over a wide range (10.2~30.5 h) by changing the mix ratio o grout, with NO. 7 having the lowest setting time (10.2 h) and NO. 4 having the highes (30.5 h).The setting time of NO. 1~NO.3, NO.7~NO. 10 and NO.12~NO.15 was within 10~24 h. Figure 7 illustrates the influence of five factors on setting time.The range analysi result presented that the significance of each factor on the setting time was ranked in the order of RC > RE > RB > RA = RD.It showed that muck content and water-solid ratio had a great influence on the setting time of the SMGM (RC = 9.8, RE = 7.58).The setting time of th SMGM increased with the increase in the water-solid ratio, admixture content, and fly ash content.The admixture contains the retarder.It inhibits cement hydration.Large amounts o free water were filled inside the SMGM with the increase in the water-solid ratio, which mad it difficult for the hydrated calcium silicate (C-S-H) gel formed by the cement hydration reac tion to aggregate.Therefore, the setting time of the SMGM was prolonged.This was consisten with the results of G. Zheng's study [37].The setting time of the SMGM generally decreased with an increase in muck and cement content.This is because as the content of muck and cement increases, clay particles adsorb a large number of free water molecules [38], accelerat ing the hydration reaction of cement.Additionally, the negatively charged clay particles un dergo very weak charge exchange with cement, aggregating into larger soil particles, which further shortens the setting time of the SMGM.

Influence of Factors on Setting Time
Figure 6 shows the variation in the setting time of the SMGM.The setting time of the SMGM could be adjusted over a wide range (10.2~30.5 h) by changing the mix ratio of grout, with NO. 7 having the lowest setting time (10.2 h) and NO. 4 having the highest (30.5 h).The setting time of NO. 1~NO.3, NO.7~NO. 10 and NO.12~NO.15 was within 10~24 h. Figure 7 illustrates the influence of five factors on setting time.The range analysis result presented that the significance of each factor on the setting time was ranked in the order of R C > R E > R B > R A = R D .It showed that muck content and water-solid ratio had a great influence on the setting time of the SMGM (R C = 9.8, R E = 7.58).The setting time of the SMGM increased with the increase in the water-solid ratio, admixture content, and fly ash content.The admixture contains the retarder.It inhibits cement hydration.Large amounts of free water were filled inside the SMGM with the increase in the water-solid ratio, which made it difficult for the hydrated calcium silicate (C-S-H) gel formed by the cement hydration reaction to aggregate.Therefore, the setting time of the SMGM was prolonged.This was consistent with the results of G. Zheng's study [37].The setting time of the SMGM generally decreased with an increase in muck and cement content.This is because as the content of muck and cement increases, clay particles adsorb a large number of free water molecules [38], accelerating the hydration reaction of cement.Additionally, the negatively charged clay particles undergo very weak charge exchange with cement, aggregating into larger soil particles, which further shortens the setting time of the SMGM.

Influence of Factors on Flowability
Figure 8 shows the variation in the initial flowability of the SMGM.The initial flowability of the SMGM ranged from 115 mm to 345 mm, with NO. 10 having the lowest initial flowability (115 mm) and NO.11 having the highest (345 mm). Figure 9 presents the flowability loss of the SMGM over time.The flowability of the SMGM decreased as time progressed because the number of free water molecules inside the slurry decreased and the C-S-H generated by cement hydration continuously aggregated, thereby gradually reducing the flowability of the SMGM.After 3 h from mixing, the flowability of NO. 1~NO.6, NO. 9, NO.11, NO. 12, and NO.14~NO.16 was greater than 160 mm., R E = 95.3)were much higher than R B , R A , and R D , indicating that the flowability of the SMGM was mainly affected by muck content and water-solid ratio.The flowability of the SMGM decreased with the increase in muck content.This is because soil particles have a large specific surface area and a strong ability to adsorb water molecules, so an increase in muck content leads to a reduction in the number of free water molecules, thereby decreasing the flowability [38].The flowability of the SMGM increased with the increase in the water-solid ratio because as the content of free water molecules rose, the friction between particles decreased, resulting in higher flowability of the slurry [27].The flowability of the SMGM decreased with the increase in fly ash content.When the cement content was less than 8%, the flowability of the SMGM decreased with the increase in cement content.However, when the cement content was greater than 8%, the flowability of the SMGM increased with the increase in cement content.This may be because when the cement content is less than 8%, the increase in cement content leads to more hydration products within the slurry, causing it to become thicker and reducing its flowability.However, when the cement content exceeds 8%, the unhydrated cement particles act as lubricants, slightly increasing the flowability.The impact of admixture on the flowability of the SMGM was negligible.Figure 10 illustrates the influence of five factors on flowability.The range result presented that the significance of each factor on the flowability was ranke order of RC > RE > RB > RA > RD.RC and RE (RC = 155.8,RE = 95.3)were much higher RA, and RD, indicating that the flowability of the SMGM was mainly affected b content and water-solid ratio.The flowability of the SMGM decreased with the in muck content.This is because soil particles have a large specific surface are strong ability to adsorb water molecules, so an increase in muck content leads to tion in the number of free water molecules, thereby decreasing the flowability [ flowability of the SMGM increased with the increase in the water-solid ratio be the content of free water molecules rose, the friction between particles decreased, r in higher flowability of the slurry [27].The flowability of the SMGM decreased increase in fly ash content.When the cement content was less than 8%, the flowa the SMGM decreased with the increase in cement content.However, when the content was greater than 8%, the flowability of the SMGM increased with the inc cement content.This may be because when the cement content is less than 8%, the in cement content leads to more hydration products within the slurry, causing it to thicker and reducing its flowability.However, when the cement content exceeds unhydrated cement particles act as lubricants, slightly increasing the flowability.pact of admixture on the flowability of the SMGM was negligible.

Influence of Factors on Volume Shrinkage Rate
Figure 11 shows the variation in the volume shrinkage rate of the SMGM.The shrinkage rate of the SMGM could be adjusted over a wide range (0.14~9.34%) by c the mix ratio of grout, with NO. 9 and NO. 15 having the lowest volume shrinkage rat and NO.1 having the highest (9.34%).The volume shrinkage rates of NO. 2~NO.4, NO

Mechanical Properties 3.2.1. Influence of Factors on Volume Shrinkage Rate
Figure 11 shows the variation in the volume shrinkage rate of the SMGM.The volume shrinkage rate of the SMGM could be adjusted over a wide range (0.14~9.34%) by changing the mix ratio of grout, with NO. 9 and NO. 15 having the lowest volume shrinkage rate (0.14%) and NO.1 having the highest (9.34%).The volume shrinkage rates of NO. 2~NO.4, NO.7~NO.10, and NO.12~NO.16 were below 5%. Figure 12 illustrates the influence of five factors on volume shrinkage rate.The range analysis result presented that the significance of each factor on the volume shrinkage rate was ranked in the order of R C > R E > R A > R B > R D .R C (R C = 6.14) was much higher than R E , R B , R A , and R D , indicating that the muck has a great influence on the volume shrinkage of the SMGM.The volume shrinkage rate of the SMGM decreased with the increase in the content of muck, cement, and fly ash.When the admixture content and water-solid ratio were less than 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM decreased with the increase in these two factors.However, when the admixture content and water-solid ratio exceed 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM increased with the increase in these two factors.
was much higher than RE, RB, RA, and RD, indicating that the muck has a great influence on the volume shrinkage of the SMGM.The volume shrinkage rate of the SMGM decreased with the increase in the content of muck, cement, and fly ash.When the admixture content and watersolid ratio were less than 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM decreased with the increase in these two factors.However, when the admixture content and water-solid ratio exceed 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM increased with the increase in these two factors.

Influence of Factors on Compressive Strength
Figure 13 shows the variation in the compressive strength of the SMGM with curing age.The compressive strength of the SMGM at 1 d, 3 d, 7 d, and 28 d ranged from 0.05 to 0.16 MPa, 0.1 to 0.26 MPa, 0.15 to 0.73 MPa, and 0.25 to 2.46 MPa, respectively, indicating that the compressive strength of the SMGM gradually increased with the curing age [39].At the same curing age (such as 28 d), the compressive strength of the SMGM increased with the cement content [40].The compressive strength of NO. 10 and NO.13~NO.16 was greater than 0.15 MPa at 3 d and greater than 1 MPa at 28 d. was much higher than RE, RB, RA, and RD, indicating that the muck has a great influence on the volume shrinkage of the SMGM.The volume shrinkage rate of the SMGM decreased with the increase in the content of muck, cement, and fly ash.When the admixture content and watersolid ratio were less than 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM decreased with the increase in these two factors.However, when the admixture content and water-solid ratio exceed 0.1 and 0.425, respectively, the volume shrinkage rate of the SMGM increased with the increase in these two factors.

Influence of Factors on Compressive Strength
Figure 13 shows the variation in the compressive strength of the SMGM with curing age.The compressive strength of the SMGM at 1 d, 3 d, 7 d, and 28 d ranged from 0.05 to 0.16 MPa, 0.1 to 0.26 MPa, 0.15 to 0.73 MPa, and 0.25 to 2.46 MPa, respectively, indicating that the compressive strength of the SMGM gradually increased with the curing age [39].At the same curing age (such as 28 d), the compressive strength of the SMGM increased with the cement content [40].The compressive strength of NO. 10 and NO.13~NO.16 was greater than 0.15 MPa at 3 d and greater than 1 MPa at 28 d.

Influence of Factors on Compressive Strength
Figure 13 shows the variation in the compressive strength of the SMGM with curing age.The compressive strength of the SMGM at 1 d, 3 d, 7 d, and 28 d ranged from 0.05 to 0.16 MPa, 0.1 to 0.26 MPa, 0.15 to 0.73 MPa, and 0.25 to 2.46 MPa, respectively, indicating that the compressive strength of the SMGM gradually increased with the curing age [39].At the same curing age (such as 28 d), the compressive strength of the SMGM increased with the cement content [40].The compressive strength of NO. 10 and NO.13~NO.16 was greater than 0.15 MPa at 3 d and greater than 1 MPa at 28 d.The compressive strength of the SMGM decreased with the increase in the content of muck and the water-solid ratio [41][42][43].When the admixture content was less than 0.2%, the compressive strength of the SMGM increased with the increase in admixture content.However, when the admixture content was greater than 0.2%, the compressive strength of the SMGM decreased with the increase in admixture content.Figure 14 illustrates the influence of five factors on the compressive strength of the SMGM at different curing ages.The range analysis result presented that the significance of each factor on the compressive strength was ranked in the order of RA > RE > RC > RD > RB.When the curing time was short, the range values of different factors were relatively small.As the curing time increased, the range value for cement content (RA = 1.22) became much greater than those of the other factors, indicating that cement has a significant impact on the compressive strength of the SMGM.At 28 d of curing, the compressive strength of the SMGM increased with the increase in the content of cement and fly ash.This is because the hydration reaction of cement generates C-S-H gel and Ca(OH)2.In an alkaline environment, Ca 2+ reacts with the active [SiO4] 4− and [AlO4] 5− in fly ash to form C-S-H. Additionally, Ca 2+ undergoes ion exchange reactions with Na+ and K+ ions, causing changes in the electric double layer within the soil, leading to the attraction and aggregation of soil particles into larger particles.The aggregation of soil particles and the formation of C-S-H gel enhance the density of the grouted stone body, thereby increasing its compressive strength.The compressive strength of the SMGM decreased with the increase in the content of muck and the water-solid ratio [41][42][43].When the admixture content was less than 0.2%, the compressive strength of the SMGM increased with the increase in admixture content.However, when the admixture content was greater than 0.2%, the compressive strength of the SMGM decreased with the increase in admixture content.In order to improve the construction efficiency and ensure the effectiveness of grouting, this paper predicted the performance of the SMGM.Based on the results of 16 groups of orthogonal tests, the predictive regression model was established through multiple linear regression.The prediction model is shown in Table 5.In order to improve the construction efficiency and ensure the effectiveness of grouting, this paper predicted the performance of the SMGM.Based on the results of 16 groups of orthogonal tests, the predictive regression model was established through multiple linear regression.The prediction model is shown in Table 5. Note: F b , F f , F s , F v , F 3-c , and F 28-c were bleeding ratio, initial flowability, setting time, volume shrinkage rate, 3 d, and 28 d compressive strength, respectively.X 1 , X 2 , X 3 , and X 4 represented the content of cement, fly ash, muck, and admixture, respectively.X 5 represented the water-solid ratio.
It can be seen from Table 5 that the fitted linear regression coefficients are all greater than 0.8, and the significance levels are all less than 0.05, indicating that the model is highly statistically significant.This can provide a reference for the practical application of the SMGM.
In summary, NO. 14 and NO. 15 meet the performance requirements of grouting slurry.The performance indicators of NO. 14 and NO. 15 are shown in Figure 15.To simplify the construction process (since NO. 15 requires additional admixture), NO. 14 can be selected as the mix proportion for synchronous grouting.

Microstructure Properties
To explore the internal structure of the SMGM and analyze the intrinsic relationship between their strength and microstructure, XRD, SEM, and CT tests were conducted on NO. 4 and NO. 14 samples cured for 28 d.

XRD Analysis
XRD is a rapid method for detecting the phase composition of materials.The height of the diffraction peaks reflects the content of the phase components.The higher the peak, the greater the content.The XRD diffraction results of the SMGM are shown in Figure 16.The hardened SMGM contained SiO2 and CaAl2Si2O8•4H2O, originating from unreacted solid particles; AFt from the hydration products of cement; and CaCO3 from the chemical reaction between carbonates in the soil and the hydration product Ca(OH)2.The cement hydration reaction generated C-S-H, AFt, and Ca(OH)2 [44,45].Part of the Ca(OH)2 reacted with carbonates in the soil to form CaCO3, while another part underwent pozzolanic reactions with SiO2 and Al2O3 in FA [46].Compared with NO. 4, NO. 14 has a higher cement content, resulting in more hydration products and greater consumption of SiO2 in subsequent reactions.Therefore, NO. 14 exhibits lower SiO2 diffraction peak values and higher AFt and CaCO3 diffraction peak values than NO. 4.

Microstructure Properties
To explore the internal structure of the SMGM and analyze the intrinsic relationship between their strength and microstructure, XRD, SEM, and CT tests were conducted on NO. 4 and NO. 14 samples cured for 28 d.

XRD Analysis
XRD is a rapid method for detecting the phase composition of materials.The height of the diffraction peaks reflects the content of the phase components.The higher the peak, the greater the content.The XRD diffraction results of the SMGM are shown in Figure 16.The hardened SMGM contained SiO 2 and CaAl 2 Si 2 O 8 •4H 2 O, originating from unreacted solid particles; AFt from the hydration products of cement; and CaCO 3 from the chemical reaction between carbonates in the soil and the hydration product Ca(OH) 2 .The cement hydration reaction generated C-S-H, AFt, and Ca(OH) 2 [44,45].Part of the Ca(OH) 2 reacted with carbonates in the soil to form CaCO 3 , while another part underwent pozzolanic reactions with SiO 2 and Al 2 O 3 in FA [46].Compared with NO. 4, NO. 14 has a higher cement content, resulting in more hydration products and greater consumption of SiO 2 in subsequent reactions.Therefore, NO. 14 exhibits lower SiO 2 diffraction peak values and higher AFt and CaCO 3 diffraction peak values than NO. 4. reaction between carbonates in the soil and the hydration product Ca(OH)2.The cement hydration reaction generated C-S-H, AFt, and Ca(OH)2 [44,45].Part of the Ca(OH)2 reacted with carbonates in the soil to form CaCO3, while another part underwent pozzolanic reactions with SiO2 and Al2O3 in FA [46].Compared with NO. 4, NO. 14 has a higher cement content, resulting in more hydration products and greater consumption of SiO2 in subsequent reactions.Therefore, NO. 14 exhibits lower SiO2 diffraction peak values and higher AFt and CaCO3 diffraction peak values than NO. 4.

SEM Analysis
Figure 17 shows the micromorphology of NO. 4 and NO. 14 at 2 k and 8 k times.In NO. 4, there were various shapes and sizes of particles with relatively small contact areas between them, leading to a loose structure and the formation of numerous pores.Unhydrated FA was present as individual particles.In contrast, NO. 14 shows more aggregated soil particles, more

SEM Analysis
Figure 17 shows the micromorphology of NO. 4 and NO. 14 at 2 k and 8 k times.In NO. 4, there were various shapes and sizes of particles with relatively small contact areas between them, leading to a loose structure and the formation of numerous pores.Unhydrated FA was present as individual particles.In contrast, NO. 14 shows more aggregated soil particles, more gel substances, and smaller pores.This is due to the higher cement content in NO. 14, which generates more gel substances.These gel substances fill existing pores and bind the large particles aggregated through cohesion, reducing the porosity [47,48].As a result, the internal structure of the SMGM becomes denser, significantly enhancing their mechanical properties.gel substances, and smaller pores.This is due to the higher cement content in NO. 14, which generates more gel substances.These gel substances fill existing pores and bind the large particles aggregated through cohesion, reducing the porosity [47,48].As a result, the internal structure of the SMGM becomes denser, significantly enhancing their mechanical properties.

CT Analysis
Soil pores are categorized into macropores and micropores.According to Luxmoore, pores smaller than 10 µm are defined as micropores, which have a minor effect on the strength of the SMGM [49].Due to the limitation of image resolution, this study analyzes only the characteristics of macropores in SMGM based on the 3D reconstruction model.Specifically, the porosity and pore numbers of NO. 4 and NO. 14 were divided into zones, and the distribution of macropores within these zones in the specimens is shown in Figures 18 and 19.The porosity and pore number for each zone are listed in Table 6.The pore numbers for NO. 4 and NO. 14 decreased with increasing pore diameter, but their porosity does not show a clear trend.Compared to NO. 4, the porosity and pore numbers for each zone in NO. 14 were smaller, with the total porosity and total pore zone of NO. 14 being 22.66% and 42.51% of NO. 4, respectively.The reason for this is the higher cement content in NO. 14, which generates more gel substances through the cement hydration reaction and the pozzolanic effect of fly ash.These gel substances fill the internal pores.Additionally, due to particle aggregation, small particles cluster together to form larger particles.Consequently, these gel substances and soil particle aggregation reduce the pore numbers and decrease porosity [50].The proportion of porosity and pore numbers for each zone of NO. 4 and NO. 14 within the macropores is shown in Figure 20.The macropores in NO. 4 and NO. 14 were mainly concentrated in the range of 28~200 μm.The pore numbers within the 28~200 μm range account for 86.04% and 85.92% of the total macropore numbers for NO. 4 and NO.14, respectively.Compared to NO. 4, NO. 14 had an increased porosity in the 28~400 μm range and a significantly reduced porosity for pores larger than 800 μm.The proportion of porosity and pore numbers for each zone of NO. 4 and NO. 14 within the macropores is shown in Figure 20.The macropores in NO. 4 and NO. 14 were mainly concentrated in the range of 28~200 μm.The pore numbers within the 28~200 μm range account for 86.04% and 85.92% of the total macropore numbers for NO. 4 and NO.14, respectively.Compared to NO. 4, NO. 14 had an increased porosity in the 28~400 μm range and a significantly reduced porosity for pores larger than 800 μm.The proportion of porosity and pore numbers for each zone of NO. 4 and NO. 14 within the macropores is shown in Figure 20.The macropores in NO. 4 and NO. 14 were mainly concentrated in the range of 28~200 µm.The pore numbers within the 28~200 µm range account for 86.04% and 85.92% of the total macropore numbers for NO. 4 and NO.14, respectively.Compared to NO. 4, NO. 14 had an increased porosity in the 28~400 µm range and a significantly reduced porosity for pores larger than 800 µm.In conclusion, compared to NO. 4, NO. 14 has a higher amount of hydration p These hydration products and particle agglomeration significantly improve the pore structure of the SMGM, reducing their porosity [51,52].Consequently, NO. 1 its higher compressive strength, consistent with the results shown in Figure 13.

Conclusions
This study attempted to utilize shield muck for synchronous grouting.Thr thogonal experiments, the influence of the water-solid ratio and the content of cem ash, muck, and admixture on the fresh properties, mechanical properties, and mic ture of the SMGM were investigated.The feasibility of shield muck as a sync grouting raw material was verified from both macro and micro perspectives.Base analysis of the experimental results, the following conclusions are drawn: (1) The range analysis results indicate that the influence of muck content and wa ratio on the bleeding ratio, flowability, setting time, and volume shrinkage rate than that of other factors.The influence of cement on the compressive strength than other factors.A predictive model for the performance of the SMGM was es by using SPSS software, providing a reference for its practical application.(2) Through the combination of macro and micro methods, it was found that a dense structure is the main reason for the higher compressive strength.The analysis that the filling effect of the gel substances and the particle aggregation effect re pores of the SMGM, significantly improving their pore structure and compactn (3) To meet the construction performance requirements and simplify the cons process, NO. 14 (12% for cement, 18% for fly ash, 50% for muck, 0.465 for wa ratio, 19.5% for river sand, 0.5% for bentonite) can be selected as the sync In conclusion, compared to NO. 4, NO. 14 has a higher amount of hydration products.These hydration products and particle agglomeration significantly improve the internal pore structure of the SMGM, reducing their porosity [51,52].Consequently, NO. 14 exhibits higher compressive strength, consistent with the results shown in Figure 13.

Conclusions
This study attempted to utilize shield muck for synchronous grouting.Through orthogonal experiments, the influence of the water-solid ratio and the content of cement, fly ash, muck, and admixture on the fresh properties, mechanical properties, and microstructure of the SMGM were investigated.The feasibility of shield muck as a synchronous grouting raw material was verified from both macro and micro perspectives.Based on the analysis of the experimental results, the following conclusions are drawn: (1) The range analysis results indicate that the influence of muck content and water-solid ratio on the bleeding ratio, flowability, setting time, and volume shrinkage rate is greater than that of other factors.The influence of cement on the compressive strength is greater than other factors.A predictive model for the performance of the SMGM was established by using SPSS software (https://stats.oarc.ucla.edu/spss/,accessed on 14 August 2024), providing a reference for its practical application.(2) Through the combination of macro and micro methods, it was found that a dense internal structure is the main reason for the higher compressive strength.The analysis suggests that the filling effect of the gel substances and the particle aggregation effect reduce the pores of the SMGM, significantly improving their pore structure and compactness.(3) To meet the construction performance requirements and simplify the construction process, NO. 14 (12% for cement, 18% for fly ash, 50% for muck, 0.465 for water-solid ratio, 19.5% for river sand, 0.5% for bentonite) can be selected as the synchronous

Figure 1 .
Figure 1.The particle size distribution of sand, bentonite, and loess.

Figure 1 .
Figure 1.The particle size distribution of sand, bentonite, and loess.

Figure 1 .
Figure 1.The particle size distribution of sand, bentonite, and loess.

Figure 4 .
Figure 4. Variation in bleeding ratio of the SMGM.

Figure 4 .
Figure 4. Variation in bleeding ratio of the SMGM.

Figure 4 .
Figure 4. Variation in bleeding ratio of the SMGM.

Figure 5 .
Figure 5. Range analysis of bleeding ratio.

Figure 5 .
Figure 5. Range analysis of bleeding ratio.

Figure 6 .
Figure 6.Variation in setting time of the SMGM.Figure 6. Variation in setting time of the SMGM.

Figure 6 .
Figure 6.Variation in setting time of the SMGM.Figure 6. Variation in setting time of the SMGM.

Figure 6 .
Figure 6.Variation in setting time of the SMGM.

Figure 7 .
Figure 7. Range analysis of setting time.

Figure 8 .
Figure 8. Variation in initial flowability of the SMGM.

Figure 7 .
Figure 7. Range analysis of setting time.3.1.3.Influence of Factors on FlowabilityFigure8shows the variation in the initial flowability of the SMGM.The initial flowability of the SMGM ranged from 115 mm to 345 mm, with NO. 10 having the lowest initial flowability (115 mm) and NO.11 having the highest (345 mm).Figure9presents the flowability loss of the SMGM over time.The flowability of the SMGM decreased as time progressed because the number of free water molecules inside the slurry decreased and the C-S-H generated by cement hydration continuously aggregated, thereby gradually reducing the flowability of the SMGM.After 3 h from mixing, the flowability of NO. 1~NO.6, NO. 9, NO.11, NO. 12, and NO.14~NO.16 was greater than 160 mm.
Figure8shows the variation in the initial flowability of the SMGM.The initial flowability of the SMGM ranged from 115 mm to 345 mm, with NO. 10 having the lowest initial flowability (115 mm) and NO.11 having the highest (345 mm).Figure9presents the flowability loss of the SMGM over time.The flowability of the SMGM decreased as time progressed because the number of free water molecules inside the slurry decreased and the C-S-H generated by cement hydration continuously aggregated, thereby gradually reducing the flowability of the SMGM.After 3 h from mixing, the flowability of NO. 1~NO.6, NO. 9, NO.11, NO. 12, and NO.14~NO.16 was greater than 160 mm.

Figure 6 .
Figure 6.Variation in setting time of the SMGM.

Figure 7 .
Figure 7. Range analysis of setting time.3.1.3.Influence of Factors on FlowabilityFigure8shows the variation in the initial flowability of the SMGM.The initial flowability of the SMGM ranged from 115 mm to 345 mm, with NO. 10 having the lowest initial flowability (115 mm) and NO.11 having the highest (345 mm).Figure9presents the flowability loss of the SMGM over time.The flowability of the SMGM decreased as time progressed because the number of free water molecules inside the slurry decreased and the C-S-H generated by cement hydration continuously aggregated, thereby gradually reducing the flowability of the SMGM.After 3 h from mixing, the flowability of NO. 1~NO.6, NO. 9, NO.11, NO. 12, and NO.14~NO.16 was greater than 160 mm.
Figure8shows the variation in the initial flowability of the SMGM.The initial flowability of the SMGM ranged from 115 mm to 345 mm, with NO. 10 having the lowest initial flowability (115 mm) and NO.11 having the highest (345 mm).Figure9presents the flowability loss of the SMGM over time.The flowability of the SMGM decreased as time progressed because the number of free water molecules inside the slurry decreased and the C-S-H generated by cement hydration continuously aggregated, thereby gradually reducing the flowability of the SMGM.After 3 h from mixing, the flowability of NO. 1~NO.6, NO. 9, NO.11, NO. 12, and NO.14~NO.16 was greater than 160 mm.

Figure 8 .
Figure 8. Variation in initial flowability of the SMGM.Figure 8. Variation in initial flowability of the SMGM.

Figure 10
Figure 10 illustrates the influence of five factors on flowability.The range analysis result presented that the significance of each factor on the flowability was ranked in the order of RC > RE > RB > RA > RD.RC and RE (RC = 155.8,RE = 95.3)were much higher than RB, RA, and RD, indicating that the flowability of the SMGM was mainly affected by muck

Figure 10
Figure10illustrates the influence of five factors on flowability.The range analysis result presented that the significance of each factor on the flowability was ranked in the order of R C > R E > R B > R A > R D .R C and R E (R C = 155.8,R E = 95.3)were much higher than R B , R A , and R D , indicating that the flowability of the SMGM was mainly affected by muck content and water-solid ratio.The flowability of the SMGM decreased with the increase in muck content.This is because soil particles have a large specific surface area and a strong ability to adsorb water molecules, so an increase in muck content leads to a reduction in the number of free water molecules, thereby decreasing the flowability[38].The flowability of the SMGM increased with the increase in the water-solid ratio because as the content of free water molecules rose, the friction between particles decreased, resulting in higher flowability of the slurry[27].The flowability of the SMGM decreased with the increase in fly ash content.When the cement content was less than 8%, the flowability of the SMGM decreased with the increase in cement content.However, when the cement content was greater than 8%, the flowability of the SMGM increased with the increase in cement content.This may be because when the cement content is less than 8%, the increase in cement content leads to more hydration products within the slurry, causing it to become thicker and reducing its flowability.However, when the cement content exceeds 8%, the unhydrated cement particles act as lubricants, slightly increasing the flowability.The impact of admixture on the flowability of the SMGM was negligible.

Figure 12 .
Figure 12.Range analysis of volume shrinkage rate.

Figure 12 .
Figure 12.Range analysis of volume shrinkage rate.

Figure 12 .
Figure 12.Range analysis of volume shrinkage rate.

Figure 14
Figure 14 illustrates the influence of five factors on the compressive strength of the SMGM at different curing ages.The range analysis result presented that the significance of each factor on the compressive strength was ranked in the order of RA > RE > RC > RD > RB.When the curing time was short, the range values of different factors were relatively small.As the curing time increased, the range value for cement content (RA = 1.22) became much

Figure 13 .
Figure 13.Compressive strength of the SMGM with different curing ages.

Figure 14
Figure14illustrates the influence of five factors on the compressive strength of the SMGM at different curing ages.The range analysis result presented that the significance of each factor on the compressive strength was ranked in the order of R A > R E > R C > R D > R B .When the curing time was short, the range values of different factors were relatively small.As the curing time increased, the range value for cement content (R A = 1.22) became much greater than those of the other factors, indicating that cement has a significant impact on the compressive strength of the SMGM.At 28 d of curing, the compressive strength of the SMGM increased with the increase in the content of cement and fly ash.This is because the hydration reaction of cement generates C-S-H gel and Ca(OH) 2 .In an alkaline environment, Ca 2+ reacts with the active [SiO 4 ] 4− and [AlO 4 ] 5− in fly ash to form C-S-H. Additionally, Ca 2+ undergoes ion exchange reactions with Na+ and K+ ions, causing changes in the electric double layer within the soil, leading to the attraction and aggregation of soil particles into larger particles.The aggregation of soil particles and the formation of C-S-H gel enhance the density of the grouted stone body, thereby increasing its compressive strength.The compressive strength of the SMGM decreased with the increase in the content of muck and the water-solid ratio[41][42][43].When the admixture content was less than 0.2%, the compressive strength of the SMGM increased with the increase in admixture content.However, when the admixture content was greater than 0.2%, the compressive strength of the SMGM decreased with the increase in admixture content.

Figure 13 .
Figure 13.Compressive strength of the SMGM with different curing ages.

Figure 14 .
Figure 14.Range analysis of compressive strength of the SMGM at different curing ages: (a) 1 d curing age; (b) 3 d curing age; (c) 7 d curing age; (d) 28 d curing age.

Figure 14 .
Figure 14.Range analysis of compressive strength of the SMGM at different curing ages: (a) 1 d curing age; (b) 3 d curing age; (c) 7 d curing age; (d) 28 d curing age.

18 Figure 15 .
Figure 15.Radar graph of performance indexes of NO. 14 and NO. 15.

Figure 15 .
Figure 15.Radar graph of performance indexes of NO. 14 and NO. 15.

Figure 17 .Figure 17 .
Figure 17.SEM analysis of NO. 4 and NO.14.(a) The image of NO. 4 magnified 2000 times.(b) The image of NO.4 magnified 8000 times.(c) The image of NO. 14 magnified 2000 times.(d) The image of NO. 14 magnified 8000 times.3.3.3.CT AnalysisSoil pores are categorized into macropores and micropores.According to Luxmoore, pores smaller than 10 μm are defined as micropores, which have a minor effect on the strength of the SMGM[49].Due to the limitation of image resolution, this study analyzes only the char-

Figure 20 .
Figure 20.Pore proportion diagram of each zone.

Figure 20 .
Figure 20.Pore proportion diagram of each zone.

Table 3 .
Orthogonal experimental design table.

Table 4 .
Orthogonal test design of the SMGM.

Table 5 .
Performance prediction of the SMGM.

Table 5 .
Performance prediction of the SMGM.