Experimental Investigation into the Mechanical Performance of Roadway Shotcrete with the Partial Replacement of Cement with Recycled Gangue Powder

Abstract

To maximize the comprehensive utilization of gangue waste, broken gangue can be used to replace gravel as the coarse aggregate to prepare underground roadway shotcrete, and treated gangue powder can be used for the partial replacement of cement. This not only diminishes the demand for conventional raw materials but also increases the amount of gangue waste disposed. Broken gangue waste was ball-milled for 1 h, 3 h, and 5 h to prepare gangue powder, which was used to partially replace cement. Then, experimental schemes for the performance of shotcrete at the rates of cement replacement of 30%, 40%, and 50% were devised to compare the mineral compositions and microscopic characteristics of shotcrete with the partial replacement of cement with gangue powder. The influences of the partial replacement of cement with gangue powder on the slump, tensile strength, and compressive strength of the shotcrete were revealed. The experimental results revealed an inverse relationship between shotcrete slump and both the cement replacement ratio and the gangue Ball-milling duration. Increasing the cement replacement ratio from 30% to 50% reduced slump by 55.3% (103 mm → 46 mm), while extending the Ball-milling time from 1 h to 5 h decreased it by 33.0% (103 mm → 69 mm). Mechanical properties showed contrasting trends: After a 28-day curing process, compressive and tensile strengths declined by 54.5% (20.18 → 9.18 MPa) and 40.4% (1.56 → 0.93 MPa), respectively, with a higher cement replacement ratio. Conversely, prolonged Ball-milling duration improved the compressive strength by 12.8% (18.68 → 21.07 MPa) and the tensile strength by 34.1% (1.26 → 1.69 MPa). Moreover, the shotcrete meets the strength requirements for engineering applications only when the cement replacement ratio is 30% with gangue ball-milling durations of 3 h and 5 h. The research provides strong support for the performance optimization of gangue-based shotcrete and the improvement of the utilization of gangue waste.

1. Introduction

Coal mines produce a huge amount of gangue waste during coal mining, the discharge of which accounts for about 15% to 20% of coal production [1,2]. With regard to its disposal, gangue waste is generally directly discharged and stockpiled on the surface to form gangue dumps, which not only occupy large areas of land but also pollute the air and groundwater and even cause disasters such as landslides and explosions [3,4,5]. Therefore, improving the comprehensive utilization of gangue waste and innovating the resource utilization technologies of gangue are fundamental solutions for solving problems pertaining to gangue [6,7,8]. In mine engineering, shotcrete support of roadways has become a widely used method [9,10] and is generally used in conjunction with devices such as anchor-bolt and bolt-mesh-anchor support [11,12,13] to achieve the high-strength support of roadways [14,15]; however, huge amounts of raw materials are needed to make the shotcrete required for this technique [16,17,18]. Conventional shotcrete is generally prepared with gravel, sand, and cement [19,20,21]. Considering the similar chemical and mineral compositions of most gangue with gravel, gangue can be used as a substitute material for gravel in shotcrete [22,23], and the modified gangue powder can partially replace cement. This not only reduces the demand for conventional raw materials but also increases the amount of gangue disposed. However, the utilization of gangue powder as a cement substitute in shotcrete production has yet to be implemented, primarily due to insufficient research and the technical and economic challenges associated with costly and complex activation processes. Therefore, it is necessary to study the mechanical performance of shotcrete with cement replaced by gangue powder, in a bid to provide support for the development of novel gangue-based shotcrete.

Scholars have conducted extensive laboratory experiments to investigate the properties of shotcrete incorporating waste materials, achieving notable advancements in understanding its performance. By replacing sand with scrap glass, Mehdi et al. developed novel shotcrete with a higher strength [24]. Alcayaga et al. discussed the feasibility of using tailings as a replacement for a fine aggregate (sand) in shotcrete production. Results indicate that the coal gangue-containing shotcrete achieves comparable or even higher mechanical properties at both early-age and 28-day curing periods [25]. Hasan et al. prepared shotcrete specimens by replacing natural aggregates with ferrochrome slag aggregates and thus developed a good gamma-ray and neutron-radiation shielding material [26]. By using graded tailings to replace sandstone as the aggregate in shotcrete, Fan et al. prepared wet-mix shotcrete and significantly improved the strength of the concrete [27]. Xiao et al. developed novel shotcrete by replacing conventional aggregates with broken inactive coal gangue, attempting to mitigate environmental problems pertinent to the mining and depletion of raw materials for producing concrete and the disposal of gangue [28]. Hwalla et al. investigated the sprayability and mechanical properties of geopolymer-based shotcrete prepared with varying ratios of fly ash to silica fume as a cement replacement [29]. Zou et al. developed high-performance shotcrete using solid wastes such as fly ash and identified its optimal mix proportion [30,31,32]. Zheng developed a shotcrete material using red mud modified through chemical and physical activation methods and determined its optimal replacement ratio [33]. The state of research on utilizing waste materials in concrete production is summarized in

Table 1. Summary of research advances in shotcrete production using waste materials.

Material	Replaced Materials	Replacement Ratio	Results
Scrap glass	Sand	0~100%	Enhanced low-substitution efficiency
Tailings	Sand	100%	Mechanical compliance achieved
Ferrochrome slag	Natural aggregates	50%	Enhanced mechanical–durability performance
Tailings	Natural aggregates and sand	100%	Mechanical compliance achieved
Gangue	Natural aggregates	100%	Mechanical compliance achieved
Geopolymer	Cement	100%	Spray mechanical compliance achieved
Fly ash	Cement	4~33%	Mechanical compliance achieved
Red mud	Cement	10~25%	Mechanical compliance achieved

. However, the aforementioned studies did not explore the mechanical performance of shotcrete with the replacement of cement with gangue powder. Considering this, investigating the properties of shotcrete with partial cement replacement with gangue powder in mine roadways can provide theoretical support for promoting its broader application.

In view of this, broken gangue was ball-milled for 1 h, 3 h, and 5 h to prepare gangue powder, which was used to partially replace cement. The experimental schemes for the performance of shotcrete prepared at rates of cement replacement of 30%, 40%, and 50% were designed to compare the mineral compositions and microscopic characteristics after the partial replacement of cement with gangue powder. The influences of the partial replacement of cement with gangue powder on the slump, tensile strength, and compressive strength of the shotcrete were revealed, providing a theoretical basis for optimizing the performance of the gangue-based shotcrete and improving the utilization of gangue waste.

2. Material and Methods

2.1. Materials

The experimental materials included the coarse aggregate, fine aggregate, cementitious materials, water, and an accelerator. Therein, the coarse aggregate was gangue and was collected from the Yangcheng Coal Mine (Jining City, Shandong Province, China). Limited by the internal diameter of the sprayer, the gangue used in the experiments was crushed to particles finer than 10 mm. The fine aggregate was sand, of which the fineness modulus was larger than 2.5 according to the requirement for shotcrete preparation. The cementitious materials were ordinary Portland cement (P.O 42.5), meeting the requirement in the preparation standard of shotcrete and the ball-milled gangue powder. The accelerator selected was an alkali-free accelerator.

The selected gangue and cement samples were analyzed using X-ray fluorescence spectroscopy (XRF) following standardized procedures. Specifically, the gangue material was initially ground and sieved through a 200-mesh sieve to achieve the appropriate particle size. Subsequently, the powdered gangue or cement was homogenized with boric acid and paraffin wax in predetermined ratios and then compacted into pellets. The prepared specimens were mounted onto the sample holder of the XRF spectrometer for elemental characterization. Through the measurement of characteristic X-ray wavelengths and intensities emitted by the samples, quantitative analysis was performed by converting the spectral data into elemental concentrations using instrument calibration curves. The derived major chemical compositions of gangue and cement are presented in

Table 2. Main chemical composition of gangue and cement.

Materials	Chemical Composition (wt.%)
	SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	SO3	Others
Gangue	62.58	18.81	2.64	3.27	1.78	2.86	0.33	7.73
Ordinary Portland cement	22.53	5.22	3.53	61.04	1.73	0.56	2.31	3.08

.

2.2. Specimen Preparation

The optimal ratio of raw materials for preparing gangue-based roadway shotcrete was ascertained on the basis of the results of initial trials, that is, the gangue/sand/cement/water/accelerator ratio was 44.5:44.5:25:16.5:1. On the basis of the ratio, gangue powder was used to replace 30%, 40%, and 50% of the cement. When preparing the specimens, the coarse aggregate, fine aggregate, and cement weighed in advance were firstly mixed and stirred for 5 min; then, the uniformly mixed accelerator and water were poured into the dry mixture, followed by stirring for 5 min. The stirred slurry was placed in three layers in molds measuring 100 mm × 100 mm × 100 mm. Then, the vibrating platform was used to vibrate the slurry (15~30 s); the vibration process should not only make the slurry inside the bubble completely escape but also prevent over-vibration to produce segregation. After smoothing the surface (localized depressions and bumps not exceeding 0.1 mm), the specimens were prepared. The prepared specimens were placed in a curing box for 1 d, 7 d, and 28 d at a temperature of (20 ± 2) °C and a relative humidity of 95%. Thereafter, the specimens were removed to measure the mechanical performance of the shotcrete. The specimen preparation process is illustrated in Figure 1.

2.3. Preparation of Gangue Powder

A jaw crusher was used to crush large gangue blocks, and a vibrating screen was adopted to screen gangue particles smaller than 5 mm. Then, the crushed gangue was sent to the ball mill through a feeder. After turning on the ball mill, the pressure generated due to the rotation of the grinding bowl drove the rotated mill ball to strike the inner wall of the grinding bowl, thus crushing the gangue particles. Under the friction and impact forces generated in the motion, gangue can be rapidly ground to powder; after ball-milling, the particles were inspected for uniformity and the absence of large granules. The preparation of the gangue powder is shown in Figure 2.

2.3.1. Particle Size Distribution

The particle size analysis of dry powder using laser diffraction was performed on the gangue powder samples obtained via ball-milling for 1 h, 3 h, and 5 h and the cement used in the experiments. Additionally, professional software was used to analyze the particle size. The distribution characteristics of particle sizes are displayed in Figure 3.

The average particle sizes of D₁₀, D₅₀, and D₉₀ are generally used to characterize the concentrated distribution characteristics of the particle sizes of the samples. D₁₀, D₅₀, and D₉₀ represent the particle sizes of powder specimens with the sifted weights accounting separately for 10%, 50%, and 90% of the total. Trends in various parameters are shown in Figure 4.

Analysis of Figure 4 shows that after ball-milling the gangue, the average particle sizes of D₁₀, D₅₀, and D₉₀ all decrease apace. As the ball-milling duration extends, the average particle sizes of D₁₀, D₅₀, and D₉₀ separately reduce from 64.765 μm to 58.729 μm, from 4.963 μm to 0.901 μm, from 26.352 μm to 12.612 μm, and from 171.116 μm to 138.948 μm, indicative of a significant grinding effect. After being ball-milled, the gangue powder shows a wider distribution range of particle sizes. This is mainly because the ball mill fails to grind the gangue uniformly, leading to a large difference between the minimum and maximum particle sizes. After ball milling, the average particle size and D₁₀ of gangue are both larger, with D₁₀ and D₅₀ being smaller, than that of cement. This indicates that the ball-milled gangue contains more small-size particles. As the ball-milling duration prolongs, the particle size of gangue gradually reduces while the reduction itself tends to stabilize.

2.3.2. Mineral Compositions of Gangue

During the ball-milling of broken gangue, mineral compositions in gangue are subjected to the externally applied mechanical force, so the crystal structure and micro-morphology may change. This process is likely to induce the transformation of the mineral compositions in the gangue, thus influencing the properties. An intelligent multi-functional X-ray diffractometer (XRD) (Smart Lab SE, Rigaku, Tokyo, Japan) was employed to measure the mineral compositions of cement and gangue powder attained via ball-milling for 1 h, 3 h, and 5 h. The XRD spectra of cement and gangue powder specimens are shown in Figure 5.

Analysis of Figure 5 reveals that the cement and gangue powder have different mineral compositions. The cement mainly comprises calcium silicate oxide, larnite, and gypsum, while gangue powder specimens mainly include quartz, kaolinite, calcite, and muscovite. As the ball-milling duration is increased, quartz, the main mineral composition of gangue, shows a slight variation in the diffraction intensity at 21°, 27°, and 50.5°; the diffraction intensities of kaolinite and calcite in gangue gradually attenuate at 12.5°, 25°, and 28°. Since kaolinite and calcite begin to decompose at temperatures above 400 °C and 897 °C, respectively, and the ball milling process typically operates below 90 °C, this indicates that their structural degradation under mechanical force is not thermally induced but rather caused by applied stress.

2.3.3. Micro-Morphologies of Gangue

A scanning electron microscope (SEM) (SHIMADZU, Kyoto, Japan) was used to observe and analyze the grain sizes, morphologies, and dispersed states of the cement and the ball-milled gangue powder. The sample preparation procedure was conducted as follows: A conductive carbon tape was employed as the substrate to minimize charge accumulation. The tape surface was initially cleaned with absolute ethanol, followed by the deposition of a small amount of cement powder or gangue ball-milled powders with varying milling durations. Loose particles were subsequently removed using a rubber bulb blower. Finally, a 5–10 nm carbon layer was uniformly deposited on the sample surface via sputter coating to enhance its conductivity and imaging quality. The test results are shown in Figure 6.

According to the analysis of Figure 6, cement particles show clearer edges and irregular polygons compared to the ball-milled gangue powder. Additionally, cement particles exhibit a large difference in the particle size and slight inter-adsorption of particles. With the extension of the ball-milling duration, the gangue powder is rapidly refined. The observation of the overall distribution and morphology of gangue particles under the friction, collision, and extrusion of the grinding medium shows the following changes: the particle size distribution becomes more uniform. Edges and corners are gradually worn off; edges blur, and the grains become finer. As the ball-milling duration prolongs, the aggregation of gangue particles becomes more significant. This phenomenon occurs because a high surface energy is accumulated on the particle surface in the ball-milling process, which facilitates the inter-adsorption of particles. As the ball-milling duration is extended, the adsorption becomes increasingly prominent, particularly on the surface of large particles, on which many fine particles are adsorbed.

2.4. Experimental Schemes

Since the current cement substitution rate of shotcrete is generally controlled within 30%, this experiment is conducted to improve the utilization rate of gangue and reduce the cost of cement. By setting the rates of cement replacement of gangue powder to be 30%, 40%, and 50%, experimental schemes under different ball-milling durations (1 h, 3 h, and 5 h) and different rates of cement replacement were designed. Thus, changes in the slump, compressive strength, and tensile strength were analyzed. A total of nine experiments were conducted. The specific experimental schemes are listed in

Table 3. Experimental schemes for partial replacement of cement with gangue powder.

No.	Ball-Milling Duration (for Gangue)/h	Rate of Cement Replacement/%
A1	1	30
A2	1	40
A3	1	50
B1	3	30
B2	3	40
B3	3	50
C1	5	30
C2	5	40
C3	5	50

; the proportions of the raw materials of the shotcrete are listed in

Table 4. Proportions of raw materials of the shotcrete.

No.	Ball-Milling Duration (for Gangue)/h	Gangue/kg·m−3	Sand/kg·m−3	Gangue Powder/kg·m−3	Cement/kg·m−3	Water/kg·m−3	Accelerator/kg·m−3
A1	1	801.25	801.25	135	315	247.5	18
A2	1	801.25	801.25	180	270	247.5	18
A3	1	801.25	801.25	225	225	247.5	18
B1	3	801.25	801.25	135	315	247.5	18
B2	3	801.25	801.25	180	270	247.5	18
B3	3	801.25	801.25	225	225	247.5	18
C1	5	801.25	801.25	135	315	247.5	18
C2	5	801.25	801.25	180	270	247.5	18
C3	5	801.25	801.25	225	225	247.5	18

.

3. Experimental Results and Discussion

3.1. Changes in the Slump

The evaluation of changes in the shotcrete under different ball-milling durations and rates of cement replacement is conducted to explore the influences of the two on the conveying performance of shotcrete. This is critical for determining the appropriate ball-milling duration and rate of cement replacement. Measuring the slump of shotcrete under different conditions provides an important basis for optimizing the proportions of raw materials of the shotcrete, thus ensuring the optimal conveying performance and working efficiency of the shotcrete in practice. Experimental results of the slump testing of the shotcrete with the partial replacement of cement with gangue powder are shown in Figure 7.

The analysis of Figure 7 shows that the slump of the slurry decreases as the rate of cement replacement increases. The slump is large at a rate of cement replacement of 30%. Taking gangue powder ball-milled for 1 h as an example, the slump reduces by 55.3% from 103 mm to 46 mm as the rate of cement replacement rises from 30% to 50%. This is mainly because ball-milled gangue particles are generally larger and more irregular than cement particles, and they have a higher surface roughness. Due to these characteristics, the gangue particles do not fill the interstices in the shotcrete as tightly and uniformly as cement particles, so the slump of the corresponding shotcrete declines.

With the extension of ball-milling duration, the slump of the slurry reduces. The slump is large when the ball-milling duration is 1 h. For instance, at the rate of cement replacement of 30%, the slump is decreased by 33.0% from 103 mm to 69 mm as the ball-milling is extended from 1 h to 5 h. This occurs mainly because with the prolonging ball-mining duration, gangue particles are gradually refined after being treated by the ball mill, and some particles are aggregated. The refined particles and the aggregates are poorly dispersed in the shotcrete slurry, so the slump of the slurry decreases.

3.2. Evolution of the Compressive Strength

The uniaxial compressive strength of shotcrete refers to the ratio of the maximum failure load borne by standard specimens to the stressed area under uniaxial compression. In the experiments, a WAW-1000 testing machine was used to apply load to shotcrete specimens in the displacement-controlled mode. After tests, Equation (1) was used to calculate the compressive strength of the specimen.

$$R_C={\displaystyle \frac{P}{S}}\times {10}^{-6}$$

(1)

where R_c is the compressive strength of the shotcrete specimens (MPa); P is the failure load of shotcrete specimens (N); S is the initial stressed area (m²).

For each test group, three specimens were evaluated, and the mean compressive strength was calculated from their averaged results. The test results of the compressive strengths of the shotcrete specimens with the partial replacement of cement with gangue powder are displayed in Figure 8.

The analysis of Figure 8 reveals that relatively low early-stage compressive strength and high long-term compressive strength of the specimens are observed. As the ball-milling duration of gangue extends, the compressive strength of specimens is gradually improved. The compressive strength is low when the ball-milling duration is 1 h. For example, at the rate of cement replacement of 30%, as the ball-milling duration is prolonged from 1 h to 5 h, the compressive strength of shotcrete cured for 28 d increases by 12.8% from 18.68 MPa to 21.07 MPa; that of shotcrete for 7 d grows by 41.5% from 10.16 MPa to 14.38 MPa. The increment of compressive strength of specimens is because ball-milling can improve the structure of gangue particles, refine these particles, and increase the content of active components such as free SiO₂ and Al₂O₃. After these active components are mixed with cement, minerals in cement begin to be hydrated to produce Ca(OH)₂, which undergoes secondary reactions with SiO₂ and Al₂O₃, thus forming calcium silicate hydrate (C-S-H) and calcium aluminate hydrate (C-A-H) gels. These reactions are conducive to the formation of tighter structures, thereby improving the long-term compressive strength of shotcrete. However, the addition of gangue powder dilutes cement particle concentration, slowing early-stage heat release and compromising initial strength development.

With the increase in the rate of cement replacement, the compressive strength of specimens gradually declines. The compressive strength is large at a rate of cement replacement of 30%. For instance, when the ball-milling duration is 3 h and the rate of cement replacement enlarges from 30% to 50%, the compressive strength of shotcrete cured for 28 d declines by 54.5% from 20.18 MPa to 9.18 MPa; it reduces by 28.0% from 12.72 MPa to 9.16 MPa for shotcrete cured for 7 d. The main cause of the reduction in the compressive strength is that as the proportion of cement replaced by gangue powder increases, the cement content in the shotcrete decreases accordingly. The ball-milled gangue particles are poorly cemented, which leads to an insufficiency of the cementitious material in the shotcrete, thus reducing its compressive strength.

3.3. Changes in the Tensile Strength

The tensile strength of shotcrete refers to the maximum tensile stress that can be borne by specimens at failure under uniaxial tensile stress. The tensile strength is generally evaluated through Brazilian splitting tests. In the method, a solid cylindrical specimen was placed in the testing machine, which applied the radial compressive load until fracturing of the specimen. In this way, the maximum tensile stress was ascertained. Data were computed after each experiment, and the tensile strength of specimens was calculated using Equation (2):

$$R_{\mathrm{t}}={\displaystyle \frac{2P}{\pi DL}}$$

(2)

where R_t is the tensile strength of the shotcrete specimens (MPa); P is the failure load of the shotcrete specimens (N); D is the diameter (mm); L is the thickness (mm). The average tensile strength of the specimens in each group was computed using Equation (3)

$$R_{\mathrm{t}s}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n{R}_{\mathrm{t}i}}$$

(3)

where R_ts is the average tensile strength of the shotcrete specimens (MPa); R_ti is the tensile strength of the ith shotcrete specimen (MPa); n is the number of specimens.

For each test group, three specimens were evaluated, and the mean tensile strength was calculated from their averaged results. Test results pertaining to the tensile strength of shotcrete with the partial replacement of cement with gangue powder are illustrated in Figure 9.

According to the analysis of Figure 9, the tensile strength of specimens is gradually improved with the prolongation of the ball-milling of gangue. The tensile strength is low when the ball-milling duration is 1 h. Taking shotcrete with a rate of cement replacement of 30% as an example, when the ball-milling duration is extended from 1 h to 5 h, the tensile strength of the shotcrete cured for 28 d is improved by 34.1% from 1.26 MPa to 1.69 MPa; that of the shotcrete cured for 7 d increases by 35.4% from 0.82 MPa to 1.11 MPa. The tensile strength of shotcrete is improved mainly because the prolonging ball-milling duration of gangue promotes the further activation and dispersion of gangue particles. As a result, more surfaces of gangue particles are exposed to the cement matrix, which enhances the combination of gangue particles and the cement matrix, thereby improving the overall tensile strength of the shotcrete.

With the increase in the rate of cement replacement, the tensile strength of specimens gradually declines. The shotcrete exhibits a high tensile strength when the rate of cement replacement is 30%. For instance, when the ball-milling duration is 3 h and the rate of cement replacement grows from 30% to 50%, the tensile strength of shotcrete cured for 28 d declines by 40.4% from 1.56 MPa to 0.93 MPa; for the shotcrete cured for 7 d, its tensile strength reduces by 61.9% from 1.13 MPa to 0.43 MPa. The reduction in the tensile strength of shotcrete is mainly because as the rate of cement replacement is enlarged, the cement content in shotcrete decreases, leading to a reduction in cementitious materials in the shotcrete. Considering that cement is the main cementitious material in shotcrete and the ball-milled gangue exhibits a poorer cementation effect, the shotcrete has a low tensile strength overall.

To meet practical engineering requirements, the shotcrete cured for 28 d needs to achieve compressive strengths exceeding 20 MPa and a tensile strength above 1.5 MPa. The shotcrete meets the strength requirements only when the cement replacement ratio is 30% with gangue ball-milling durations of 3 h and 5 h. Under these conditions, the compressive strengths are 20.18 MPa and 21.07 MPa, and tensile strengths reach 1.56 MPa and 1.69 MPa, respectively.

3.4. Mineral Compositions and Micro-Morphologies

3.4.1. Mineral Compositions of Specimen

Software Jade6.0 was used to analyze the mineral compositions of shotcrete without the replacement of cement and with the partial replacement of cement with gangue powder. The results are illustrated in Figure 10 and Figure 11.

The analysis of Figure 10 and Figure 11 shows that quartz, calcite, and kaolinite account for large proportions (57.2%, 23.4%, and 12.3%, respectively) of the mineral composition of shotcrete without the replacement of cement, and the proportions of other minerals including muscovite and gypsum are separately 5.6% and 1.5%. In shotcrete with the partial replacement of cement with gangue powder, minerals such as quartz, calcite, and kaolinite separately account for 59.3%, 18.7%, and 14.1%, while the proportions of other minerals including muscovite and gypsum account for 7% and 0.9%. These proportions imply that the two types of shotcrete contain similar types of minerals. The difference is that the shotcrete with the partial replacement of cement with gangue powder contains more clay minerals such as kaolinite than the gangue-based shotcrete, which results in a lower strength. However, shotcrete partially replacing cement with gangue powder contains higher proportions of clay minerals such as kaolinite. The water absorption and swelling tendency of kaolinite may induce microcracks within the shotcrete, compromising its strength. Additionally, the flaky-structured muscovite in shotcrete tends to form weak interfaces, reducing the bonding strength between aggregates and cementitious materials, thereby further lowering its mechanical performance.

3.4.2. Micro-Morphologies of Specimen

SEM was adopted to observe the surface and cross section of shotcrete specimens to obtain the surface morphology, particle distribution, and particle connection of shotcrete specimens. The micro-morphologies of shotcrete without the replacement of cement and with the partial replacement of cement with gangue powder were compared. Test results are shown in Figure 12 and Figure 13.

The analysis of Figure 12 and Figure 13 shows that the two types of shotcrete both have rough surfaces, pores, and fractures. In relative terms, the pores and fractures are larger in shotcrete with the partial replacement of cement with gangue powder. This is attributed to the lower hydration reactivity of coal gangue powder compared to cement, despite ball-milling activation. The reduced hydration products lead to higher porosity and unfilled cracks in shotcrete. These pores and fractures induce localized stress concentrations and diminish the effective load-bearing area, causing premature failure upon reaching theoretical strength thresholds. Therefore, the mechanical performance of shotcrete without the replacement of cement is superior to that of shotcrete with the partial replacement of cement with gangue powder.

4. Conclusions

Experimental schemes for the performance of shotcrete prepared at different rates of cement replacement were designed. The mineral compositions and microscopic characteristics of the shotcrete with the partial replacement of cement with gangue powder were compared. The influences of the partial replacement of cement with gangue powder on the slump, tensile strength, and compressive strength of shotcrete were unveiled. The following key conclusions were obtained:

(1)

By analyzing the experimental results of the conveying performance of shotcrete with the partial replacement of cement with gangue powder, the rate of cement replacement is found to be inversely proportional to the slump of the shotcrete. This is mainly because the ball-milled gangue particles show non-uniform shapes and particle sizes, enhanced water absorption, and raised the number of pores. The ball-milling duration of gangue is inversely proportional to the slump of shotcrete mainly because as the ball-milling duration is extended, the gangue particles are gradually refined; their specific surface area gradually increases, and particles are arranged in a less uniform manner;

(2)

The analysis of the experimental results pertaining to tests of the mechanical performance of the shotcrete with the partial replacement of cement with gangue powder reveals that the compressive and tensile strengths of the shotcrete are directly proportional to the ball-milling duration of gangue and inversely proportional to the rate of cement replacement. This is mainly because with the extension of ball-milling duration, the gangue particles are refined; the activity is improved, and the filling effect is enhanced. As the rate of cement replacement is increased, the amount of the cementitious material in the shotcrete decreases; the particles show poorer dispersion, and the porosity increases. Moreover, the shotcrete meets the strength requirements for engineering applications only when the cement replacement ratio is 30% with gangue ball-milling durations of 3 h and 5 h;

(3)

The mineral compositions and microscopic characteristics of the shotcrete with the partial replacement of cement with gangue powder are analyzed. The results show that, after partially replacing cement with gangue powder, the shotcrete contains more clay minerals such as kaolinite while having a lower cement content than the shotcrete without the replacement of cement. As a result, the content of the cement-based cementitious material reduces accordingly, leading to lower strengths of the shotcrete.

Overall, partially replacing cement with ball-milled coal gangue to produce shotcrete is feasible for underground mine roadway applications, effectively utilizing solid waste gangue and reducing cement consumption. However, studies indicate that the shotcrete with the partial replacement of cement with gangue powder exhibits lower early-age strength compared to similar materials, potentially leading to inadequate initial support strength post-spraying. Moreover, the ball-milling process for gangue is relatively complex, requiring specialized equipment and extended durations (3–5 h), posing challenges for large-scale applications in mine roadways. Future research should focus on optimizing materials and processing techniques, alongside conducting pilot projects to accumulate data for broader implementation in tunnel engineering and slope protection.