Full-hole slag concrete collaborative deformation design method

Abstract

This invention relates to the field of building materials technology, and discloses a method for the coordinated deformation design of all-cavity slag concrete. This method, based on the theory of coordinated deformation, includes five steps: determining design parameters, material selection, machine learning-based mix proportion output, trial mixing and adjustment, and determining the construction mix proportion. It establishes a raw material-performance database, combines a neural network shrinkage prediction model, a gradient-enhanced strength prediction model, and a genetic algorithm for multi-objective optimization, and outputs a recommended mix proportion. After trial mixing and adjustment to ensure workability, strength, and shrinkage rate meet standards, the construction mix proportion is determined based on the aggregate moisture content. Engineering verification shows that this method can achieve a performance balance for C30-C60 grade concrete and is suitable for the resource utilization of cavity slag aggregate.

Description

Technical Field

[0001] This invention relates to the field of building materials technology, and in particular to a method for co-deformation design of all-tunnel muck concrete. This design method is based on co-deformation theory and data-driven approach, and is applicable to the resource utilization of abandoned tunnel muck in tunnel engineering. Background Technology

[0002] Traditional concrete design prioritizes strength, neglecting the coordinated deformation of the paste, aggregate, and interface. This leads to a high risk of drying shrinkage cracking. Cement aggregates contain stone powder from various lithologies, such as granite and limestone. The dosage of these aggregates has a complex impact on shrinkage, and there is a lack of quantitative control methods based on lithological differences. Traditional concrete mix design methods rely on trial and error adjustments and have not established mathematical models of material parameters and deformation performance, making it difficult to achieve simultaneous optimization of mechanical properties and shrinkage deformation. Summary of the Invention

[0003] The purpose of this invention is to provide a collaborative deformation design method for blast furnace slag concrete, which integrates collaborative deformation theory, machine learning and multi-objective optimization to achieve balanced control of blast furnace slag concrete strength and drying shrinkage deformation.

[0004] To achieve the above objectives, the following technical solution is adopted:

[0005] This invention provides a method for the coordinated deformation design of full-cavity slag concrete, the method comprising:

[0006] S1: Determine the strength grade and workability requirements of the concrete; wherein the workability requirements include the target slump value;

[0007] S2: Cement, slag aggregate, stone powder, and water-reducing agent are used as raw materials. The cement is selected as silicate cement of appropriate grade according to the strength grade. The slag aggregate includes manufactured sand and coarse aggregate. The manufactured sand is medium sand with a fineness modulus of 2.3-3.0. The maximum particle size of the coarse aggregate is 31.5mm. The stone powder is granite stone powder or limestone stone powder. The water-reducing agent is polycarboxylate-based and has a water reduction rate of 20%-30%.

[0008] S3: Based on a pre-set database of raw materials, deformation properties, and mechanical properties, performance prediction is performed using a 56-day deformation prediction model and a 28-day mechanical prediction model. Combined with a genetic algorithm, multi-objective optimization is performed to output a recommended ratio range. The multi-objective optimization aims to maximize compressive strength and minimize shrinkage.

[0009] S4: Conduct trial mixing according to the recommended mixing ratio range, and test the slump, 7-day compressive strength and 56-day shrinkage rate. If the slump does not meet the requirements, adjust the water-reducing agent dosage or water-cement ratio; if the 7-day compressive strength is less than 70% of the design value, readjust the mixing ratio; if the 56-day shrinkage rate is not within the recommended range, readjust the mixing ratio.

[0010] S5: Detect the moisture content of manufactured sand and coarse aggregate, and adjust the amount of manufactured sand, coarse aggregate, and water according to the moisture content to obtain the construction mix proportion.

[0011] Preferably, in step S1, the strength grade of the concrete is C30-C60, and the target slump value is 100-220mm.

[0012] Preferably, in step S2, the content of the stone powder meets the following requirements: the content of non-reactive stone powder does not exceed 15%, and the content of reactive stone powder does not exceed 20%; the methylene blue value of the manufactured sand is ≤1.4, the crushing index is ≤20%, and the loose bulk density is ≥1400 kg / m³. 3 .

[0013] Preferably, in step S3, the 56d deformation prediction model is a neural network-based prediction model, and the 28d mechanical prediction model is a gradient boosting tree-based prediction model; the recommended mix ratio range includes the following 28d compressive strength range and 56d shrinkage rate range for each strength grade:

[0014] C30: 28-day compressive strength 30-40 MPa, 56-day shrinkage 351-480 × 10⁻⁶ MPa. -6 ;

[0015] C40: 28-day compressive strength 40-50 MPa, 56-day shrinkage 304-455 × 10⁻⁶ MPa. -6 ;

[0016] C50: 28-day compressive strength 50-60 MPa, 56-day shrinkage 287-439 × 10⁻⁶ MPa. -6 ;

[0017] C60: 28-day compressive strength 60-70 MPa, 56-day shrinkage 250-368 × 10⁻⁶ MPa. -6 .

[0018] Preferably, in step S3, in the recommended mix ratio range output by the multi-objective optimization, the water-cement ratio, cement dosage, sand ratio, coarse aggregate dosage, and water-reducing agent dosage are fixed values, while the dosages of fly ash, mineral powder, and stone powder are adjustable ranges; the dosage range of stone powder is determined according to the lithology, as follows:

[0019] C30: Granite powder 12.3-18.5%, limestone powder 9.5-14.8%;

[0020] C40: Granite powder 8.3-14.4%, limestone powder 6.6-11.5%;

[0021] C50: 6.3-10% granite powder, 5-8% limestone powder;

[0022] C60: Granite powder 0-6.8%, limestone powder 0-5.4%.

[0023] Preferably, step S3 further includes stone powder content correction, specifically including: using stone powder to replace fly ash and / or mineral powder by mass substitution, and determining the substitution amount based on a substitution coefficient, wherein the substitution coefficient is calculated by a formula:

[0024]

[0025] Where, μ i γ is the substitution coefficient. i σ is the correlation coefficient between stone powder and fly ash or mineral powder. S σ represents the standard deviation of stone powder content. i This represents the standard deviation of fly ash or mineral powder content.

[0026] Preferably, in step S4, the adjustment method for the slump includes:

[0027] When the slump is insufficient, increase the water-reducing agent dosage by 0.1% or fine-tune the water-cement ratio by ±0.01 each time; when the cohesiveness is poor, increase the sand ratio by 1% each time.

[0028] When controlling the gas content, if the gas content does not meet the requirement of 3%-5%, the amount of gas-entraining agent should be adjusted by 0.005% each time.

[0029] Preferably, in step S4, a cubic specimen is prepared for compressive strength testing, and a prism specimen is prepared for shrinkage testing; the cubic specimen is cured for 7 days or 28 days, and the prism specimen is cured for 3 days and then moved into a drying oven at a temperature of 20±2℃ and a humidity of 60±5% for curing.

[0030] Preferably, in step S5, the adjustment formulas for the amount of manufactured sand, coarse aggregate, and water are as follows:

[0031] Manufactured sand dosage = Theoretical manufactured sand dosage × (1 + α / 100)

[0032] Coarse aggregate usage = Theoretical coarse aggregate usage × (1 + β / 100)

[0033] Water consumption = Theoretical water consumption - Theoretical manufactured sand consumption × α / 100 - Theoretical coarse aggregate consumption × β / 100

[0034] Where α is the moisture content (%) of the manufactured sand and β is the moisture content (%) of the coarse aggregate.

[0035] Preferably, the tunnel slag aggregate is obtained by processing the tunnel slag produced during tunnel excavation, and the coarse aggregate uses stones with three particle sizes: 5-10mm, 10-20mm, and 20-31.5mm.

[0036] The beneficial effects of this invention are:

[0037] (1) The design method of slag aggregate concrete is divided into five steps: determining the design parameters (strength grade, workability), material selection (cement, machine aggregate, stone powder, admixtures, etc.); using machine learning technology to output recommended mix proportions (based on neural network drying shrinkage model, gradient improvement strength prediction model and genetic algorithm multi-objective optimization); conducting trial mix adjustments (workability, strength, shrinkage rate testing and optimization); and finally determining the construction mix proportion (considering aggregate moisture content correction).

[0038] (2) By integrating machine learning techniques (including a neural network-based drying shrinkage model and a gradient-enhanced strength prediction model) with a multi-objective genetic algorithm, a strength-drying shrinkage dual-objective optimization model was established. This model can output recommended mix proportions for C30 to C60 grade concrete to ensure that its mechanical and deformation properties are reasonably controlled.

[0039] (3) Feasibility verification in engineering application: Taking the muck aggregate of a tunnel project as an example, C40 full-muck concrete (water-cement ratio 0.48, sand ratio 45%, granite powder content 10%) was used. The measured 28-day strength was 45.1 MPa, and the 56-day drying shrinkage rate was 301 × 10⁻⁶. -6 The performance indicators all met the design requirements, verifying the engineering applicability of the collaborative design method. Attached Figure Description

[0040] Figure 1 A flowchart for the design of ordinary concrete based on existing technology;

[0041] Figure 2 A flowchart of a method for designing coordinated deformation of full-cavity slag concrete is provided in an embodiment of the present invention;

[0042] Figure 3 A flowchart of the genetic algorithm provided in an embodiment of the present invention;

[0043] Figure 4 A schematic diagram illustrating the analysis results of the shrinkage performance of concrete made from granite and limestone powder provided in this embodiment of the invention; wherein, (a) a bar chart comparing the drying shrinkage deformation of full-cavity slag concrete of different numbers at different ages; (b) a line graph showing the trend of drying shrinkage deformation of full-cavity slag concrete with age; (c) a bar chart comparing the compressive strength of full-cavity slag concrete of different numbers at different ages; and (d) a line graph showing the trend of compressive strength of full-cavity slag concrete with age. Detailed Implementation

[0044] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0045] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0046] Example 1:

[0047] like Figure 1 The diagram shown is a flowchart of the ordinary concrete design process based on existing technology. The ordinary concrete design process includes five steps: determining design parameters, material selection, preliminary mix proportion calculation, trial mixing and adjustment, and determining the construction mix proportion. The existing ordinary concrete design process has the following shortcomings:

[0048] 1) Lack of accurate performance prediction. Without utilizing technologies such as machine learning, it is difficult to accurately predict the long-term deformation (drying shrinkage) and mechanical properties of concrete. Relying on experience-based trial mixing is inefficient, costly, and difficult to adapt to the complex properties of special aggregates such as full-cavity slag.

[0049] 2) Insufficient multi-objective optimization. The system failed to systematically optimize multiple properties such as strength and deformation, relying solely on simple adjustments. This makes it difficult to achieve synergistic goals such as maximizing strength and minimizing deformation, resulting in poor synergistic control of the performance of the slag concrete throughout the tunnel.

[0050] 3) Limitations in adapting to special aggregates. For tunnel muck aggregates (including manufactured sand, stone powder, etc.), the impact of lithology and stone powder activity on performance is not considered, making it difficult to accurately design high-performance mix proportions suitable for all tunnel muck, thus limiting resource utilization.

[0051] 4) Lack of data-driven approach. Without a comprehensive raw material-performance database, it is difficult to accumulate and reuse data. When faced with new materials such as full-cavity slag, it is difficult to quickly iterate and optimize the mix ratio, resulting in a weak ability to adapt to new engineering needs.

[0052] To address the shortcomings of the existing technology, this invention provides a method for the coordinated deformation design of all-tunnel muck concrete. This design method, based on coordinated deformation theory and data-driven approaches, is applicable to the resource utilization of abandoned tunnel muck in tunnel engineering. The basic principles of the coordinated deformation theory are as follows:

[0053] Drying shrinkage is the phenomenon of concrete volume reduction during water loss, mainly caused by the hydration reaction of the paste and changes in capillary pressure. The theory of synergistic deformation design suggests that through reasonable material proportions, the deformation of the paste, aggregate, and interfaces can be coordinated, thereby reducing overall drying shrinkage. The applicant discovered a correlation between raw materials and deformation performance: the deformation performance of concrete tends to decrease with increasing water-cement ratio and sand ratio, and increases with increasing water-reducing agent dosage. Furthermore, the relationship between these three influencing factors and deformation performance is non-linear. The amount of stone powder is affected by multiple interactive factors, and the drying shrinkage value of concrete containing granite stone powder and limestone exhibits a trend of increasing and decreasing with increasing dosage. Therefore, a data-driven synergistic deformation design method for all-tunnel slag concrete is being considered.

[0054] like Figure 2 The diagram shows a flowchart of a method for co-deformation design of slag concrete in tunnels according to an embodiment of the present invention. The method includes the following steps S1-S5.

[0055] S1: Determine the design parameters.

[0056] The strength grade and workability requirements of the concrete are clearly defined. The workability requirements include the target slump value. This step S1 provides the basic parameter basis for subsequent material selection and mix design. The strength grade directly determines the cement type and mechanical property control standards in material selection, and the target slump value is the workability judgment benchmark for subsequent trial mixing and adjustment.

[0057] S2: Material selection.

[0058] Based on the strength grade determined in step S1, cement, cave debris aggregate, stone powder, and water-reducing agent are selected. Among them, ordinary Portland cement of grade 42.5 or 52.5 is selected according to the strength grade. Cave debris aggregate includes manufactured sand and coarse aggregate. The manufactured sand is medium sand with a fineness modulus of 2.3-3.0. The maximum particle size of coarse aggregate is 31.5mm. The stone powder is granite stone powder or limestone stone powder. The water-reducing agent is polycarboxylate-based with a water reduction rate of 20%-30%. The materials selected in step S2 are used as input variables for the machine learning mix proportion calculation in the subsequent step S3. The performance parameters of the materials are included in the raw material-deformation performance-mechanical property database.

[0059] S3: Machine learning output ratio.

[0060] Based on a pre-set database of raw materials, deformation properties, and mechanical properties (containing performance data of the materials selected in step S2), the performance is predicted for the strength grade determined in step S1 using a 56-day deformation prediction model and a 28-day mechanical prediction model. Combined with a genetic algorithm, the algorithm seeks to maximize compressive strength and minimize shrinkage rate through multiple objectives, and outputs a recommended mix ratio range that matches the strength grade. The mix ratio range output in this step provides an initial mix ratio basis for the trial mix in step S4, and the parameter boundaries of the mix ratio range must meet the performance requirements corresponding to workability and strength in step S1.

[0061] S4 trial fitting and adjustment.

[0062] Perform trial mixing according to the recommended mixing ratio range output in step S3, and test the slump (corresponding to the workability requirements in step S1), 7-day compressive strength (must be ≥ 70% of the strength grade design value in step S1), and 56-day shrinkage (must be within the shrinkage range recommended in step S3); if the slump does not meet the requirements, adjust the water-reducing agent dosage or water-cement ratio; if the 7-day compressive strength or 56-day shrinkage does not meet the standards, return to step S3 to re-optimize the mixing ratio range;

[0063] S5 determines the construction mix proportions.

[0064] For the mix proportion that passed the trial mix test in step S4, the moisture content of the manufactured sand and coarse aggregate selected in step S2 was tested. The amount of manufactured sand, coarse aggregate and water were adjusted according to the moisture content to obtain the construction mix proportion.

[0065] In some embodiments, the adjustment formulas for the dosage of manufactured sand, coarse aggregate, and water are as follows:

[0066] Manufactured sand dosage = theoretical manufactured sand dosage × (1 + α / 100);

[0067] Coarse aggregate usage = Theoretical coarse aggregate usage × (1 + β / 100);

[0068] Water consumption = Theoretical water consumption - Theoretical manufactured sand consumption × α / 100 - Theoretical coarse aggregate consumption × β / 100

[0069] Wherein, α is the moisture content of the manufactured sand, β is the moisture content of the coarse aggregate, and the theoretical dosage is determined based on the recommended proportion in step S3 and the adjusted proportion in step S4.

[0070] In some embodiments, in step S1, the strength grade is C30-C60, and the target slump value is 100-220mm; in step S2, the cement selection rule is: strength grade C30-C50 corresponds to 42.5 grade ordinary Portland cement, and strength grade C60 corresponds to 52.5 grade ordinary Portland cement.

[0071] In some embodiments, in step S2, the content of stone powder meets the following requirements: the content of inactive stone powder does not exceed 15%, and the content of active stone powder does not exceed 20%. In step S3, during multi-objective optimization, the content of stone powder needs to be corrected in conjunction with lithology. The content range of granite stone powder and limestone stone powder is determined according to the strength grade in step S1, specifically: C30 corresponds to 12.3-18.5% granite stone powder and 9.5-14.8% limestone stone powder; C40 corresponds to 8.3-14.4% granite stone powder and 6.6-11.5% limestone stone powder; C50 corresponds to 6.3-10% granite stone powder and 5-8% limestone stone powder; C60 corresponds to 0-6.8% granite stone powder and 0-5.4% limestone stone powder.

[0072] In some embodiments, in step S3, the 56-day deformation prediction model is a neural network-based prediction model, and the 28-day mechanical prediction model is a gradient boosting tree-based prediction model; in the multi-objective optimization output range, the water-cement ratio, cement dosage, sand ratio, coarse aggregate dosage, and water-reducing agent dosage are fixed values, while the dosages of fly ash, mineral powder, and stone powder are adjustable. Stone powder can replace part of the fly ash or mineral powder through substitution coefficients (0.32 for fly ash, 0.35 for mineral powder). The formula for calculating the substitution coefficient is:

[0073]

[0074] Where, μ i γ is the substitution coefficient. i σ is the correlation coefficient between stone powder and fly ash or mineral powder. S σ represents the standard deviation of stone powder content. i This represents the standard deviation of fly ash or mineral powder content.

[0075] In some embodiments, in step S4, cubic specimens of 150mm×150mm×150mm or 100mm×100mm×100mm are prepared for compressive strength testing, and prism specimens of 100mm×100mm×515mm are prepared for shrinkage testing. The cubic specimens are cured under standard conditions for 7 days or 28 days, and the prism specimens are cured under standard conditions for 3 days and then moved into a drying oven at a temperature of 20±2℃ and a humidity of 60±5%. The test data are used to verify whether the performance boundary of the prediction model in step S3 is met.

[0076] In some embodiments, the tunnel slag aggregate is obtained by processing the tunnel slag produced during tunnel excavation. In step S2, the coarse aggregate uses stones with three particle sizes of 5-10mm, 10-20mm, and 20-31.5mm, with a gradation of 442. In step S5, the amount of cementitious material remains unchanged, and only the amount of aggregate and water is adjusted to ensure that the water-cement ratio is consistent with the recommended ratio in step S3.

[0077] Example 2:

[0078] To further verify the engineering applicability of the co-deformation design method for full-tunnel muck concrete provided in Example 1, this example uses tunnel muck aggregate from a certain tunnel project as a case study for experimental verification. Specifically, please refer to... Figure 2 When this co-deformation design method for tunnel muck concrete was applied to the muck aggregate of a tunnel project, the design process included determining design parameters, material selection, machine learning-output mix proportions, trial mixing and adjustment, and determining the construction mix proportions. The mix proportion design method is based on the theory of co-deformation, achieving a balance between shrinkage deformation and mechanical properties through the interaction of material components. This method is based on a dual-scale design concept of microstructure control and macroscopic performance optimization, specifically including steps 1 to 6.

[0079] Step 1: Determine the design parameters.

[0080] Before using this design method, it is necessary to know the basic design objectives, determine the required strength grade and workability requirements, and select the appropriate fluidity based on the construction process. The slump of pumped concrete is generally 100-200mm. In this database, 75.0% of the data have a slump of over 180mm, and 97.3% have a slump of over 100mm. Therefore, in terms of slump, it basically meets the requirements of most pumped concrete.

[0081] Step 2: Material selection.

[0082] In this embodiment, the cement with a strength grade of C30-C50 is 42.5 grade ordinary Portland cement, and the cement fineness needs to be controlled between 300-350 μm. 2 / kg, excessively high specific surface area will exacerbate early hydration shrinkage; C60 strength grade cement is 52.5 grade ordinary Portland cement to reduce hydration heat release and shrinkage deformation. Regarding manufactured aggregates, medium sand with a fineness modulus of 2.3-3.0 is used to ensure dense particle packing; the maximum particle size of coarse aggregate is 31.5 mm, with appropriate gradation to reduce interfacial stress concentration. Regarding stone powder content, the content of non-reactive stone powder should not exceed 15%, and the content of reactive stone powder should not exceed 20%. A polycarboxylate-based water-reducing agent with a water reduction rate of 20%-30% is selected, and it must be compatible with concrete to reduce the water-cement ratio and minimize shrinkage caused by free water evaporation.

[0083] Step 3: Multi-objective mix optimization.

[0084] Step 3 achieves the transformation from "experience-based design" to "data-driven design." A database of raw materials, deformation properties, and mechanical properties is established, and a series of data analysis methods are used to statistically analyze the range of shrinkage rates for each strength grade. Shrinkage performance is assessed using a 56-day shrinkage rate model built on a neural network, while mechanical properties rely on a 28-day compressive strength prediction model built on a gradient boosting tree method. Finally, a genetic algorithm is used for multi-objective optimization (aiming to maximize compressive strength and minimize shrinkage rate) to obtain the recommended mix proportions. The process of the genetic algorithm used in this embodiment is as follows: Figure 3 As shown in Table 1, the shrinkage rate range for each strength grade is as follows.

[0085] Table 1. Range of compressive strength and shrinkage rate for each strength grade

[0086] Intensity level 28-day compressive strength range (MPa) <![CDATA[56d dry shrinkage rate range (10 -6 )]]> C30 30-40 351-480 C40 40-50 304-455 C50 50-60 287-439 C60 60-70 250-368

[0087] In the optimal solution set obtained from multi-objective optimization, some raw material parameters (such as water-cement ratio, cement dosage, sand ratio, coarse aggregate dosage, and water-reducing agent dosage) exhibit relatively concentrated fixed values ​​when different objectives are weighed. This indicates that these parameters have a strong dominant role in the objective function, and their optimal values ​​have high determinism and convergence while meeting performance requirements such as compressive strength and deformation control. Adjusting these parameters will significantly affect the performance prediction results, so the optimization algorithm tends to lock them at a relatively optimal level to ensure overall performance stability. In contrast, the values ​​of fly ash, mineral powder, and stone powder fluctuate within a certain range, thus they have a certain degree of substitutability and flexible adjustment margin when meeting performance constraints. These materials are usually used as supplementary cementitious materials or fillers, and their impact on concrete mechanics and deformation properties is relatively mild, mainly manifested in the regulation of microstructure. Therefore, the optimization process allows these parameters to vary freely within the feasible region while taking into account the objective function value, thereby improving the adaptability of mix design and the operability in actual engineering.

[0088] In the multi-objective optimization of the mix proportion of tunnel slag concrete, both mechanical and volumetric deformation properties are considered. Although silica fume is a highly reactive admixture, its variable importance in the mechanical property prediction model is not high, its impact on strength is not significant within the sample range, and it is not included in the key variables of the deformation prediction model. After comprehensive optimization and trade-offs, its contribution is limited. Stone powder, as an inert filler, has a small mechanical contribution but can improve aggregate gradation, fill pores, and increase paste density. It has good engineering adaptability in controlling drying shrinkage deformation and ensuring construction performance. Moreover, it is widely available and inexpensive, which is in line with the goals of local utilization of tunnel slag resources and sustainable development. Therefore, the optimal mix proportion scheme has a silica fume content of 0 and retains a certain proportion of stone powder. This is the result of the synergistic optimization of material properties, economy, and structural performance, reflecting the effective integration and judgment ability of the multi-objective decision-making model for complex engineering variables.

[0089] The preliminary recommended mix ratio ranges for each strength level obtained in this embodiment are shown in Table 2.

[0090] Table 2 Preliminary Recommended Mix Proportion Ranges for Each Strength Level

[0091]

[0092] Step 4: Correction of stone powder dosage.

[0093] Stone powder, as an inert filler, has the potential to improve particle size distribution and enhance the compactness of concrete. To simplify the admixture system and enhance the stability of the mixture composition, this study uses stone powder to replace fly ash and mineral powder, aiming to reduce the amount of active minerals while optimizing the construction adaptability of the mixture. In this embodiment, the correlation coefficients between stone powder and fly ash and mineral powder are 0.46 and 0.48, respectively. Based on the correlation analysis results and engineering experience, to achieve equivalent performance substitution, the following formula is used to determine the substitution coefficient of stone powder for fly ash and mineral powder.

[0094]

[0095] In the formula μ i γ is the substitution coefficient. i σ is the correlation coefficient. S σ represents the standard deviation of stone powder content. i The standard deviation of fly ash or mineral powder content is given. The calculated replacement coefficient for fly ash is 0.32, and the replacement coefficient for mineral powder is 0.35.

[0096] Table 3. Test results of drying shrinkage and compressive strength properties of concrete from the entire tunnel.

[0097]

[0098] Table 4. Test data on drying shrinkage and compressive strength of concrete at different ages.

[0099]

[0100]

[0101] according to Figure 4The analysis results of the shrinkage performance of concrete made from granite and limestone powders, as shown in Tables 3 and 4, indicate that stone powders of different lithologies have a significant impact on the shrinkage performance of concrete. The shrinkage values ​​of concrete made from granite stone powders (HG5, HG10, HG15) are significantly higher than the baseline group (DZ), and the drying shrinkage deformation gradually increases with increasing dosage. The shrinkage values ​​of concrete made from limestone stone powders (SH5, SH10, SH15) at all ages are generally lower than DZ. Although the drying shrinkage increases slightly with increasing dosage, it is still significantly lower than that of the granite group. Therefore, when recommending the dosage of stone powder, a uniform fixed range should not be adopted; instead, the differences in lithology should be considered, and specific limits should be set. Based on experimental statistical results and practical experience in materials engineering, the results are shown in Table 5.

[0102] Table 5 Recommended Partial Proportion Ranges for Each Intensity Level

[0103]

[0104] Step 5: Trial fitting and adjustment.

[0105] According to the recommended mix proportions, a forced mixer should be used, with a mixing time ≥180s to ensure uniformity. For workability indicators, the main slump is the target value, generally 100-220mm (subject to actual needs). For applications with specific air content requirements, a pressure-type air content meter should be used for testing, generally 3%-5%. If the slump is insufficient, the amount of water-reducing agent can be increased (0.1% each time) or the water-cement ratio can be finely adjusted (±0.01). If cohesiveness is poor, the sand ratio can be increased (1% each time). If the air content does not meet the requirements, the amount of air-entraining agent can be adjusted (0.005% each time) to meet the actual requirements.

[0106] After the freshly mixed concrete meets the basic requirements, prepare cubic specimens (150mm×150mm×150mm or 100mm×100mm×100mm, conversion factor 0.95) and prism specimens (100mm×100mm×515mm). Cure the cubic specimens according to standard for 7 days or 28 days and test their compressive strength. The 7-day strength should be greater than 70% of the design strength, and the 28-day strength should be greater than 100% of the design strength. After curing the prism specimens according to standard for 3 days, transfer them to a drying oven and control the temperature at 20±2℃ and the humidity at 60±5%. Measure the drying shrinkage rate at 3 days, 7 days, 14 days, 28 days, and 56 days. The shrinkage rate should be within the range of drying shrinkage rates for each strength grade in Table 5, and should also meet the principle of being as small as possible.

[0107] Step 6: Determine the construction mix proportions.

[0108] The theoretical mix design is based on dry aggregates and takes into account fluctuations in the site environment, such as changes in temperature and humidity, and aggregate mixing methods. Changes in aggregate moisture content directly affect the actual water consumption, causing the water-cement ratio to deviate from the design value. Therefore, adjusting the construction mix design can control the water-cement ratio, ensuring strength compliance and structural safety, while also reducing rework due to mix design deviations and avoiding material waste. The moisture content of the aggregates is tested. Let the moisture content of the manufactured sand be α% and the moisture content of the coarse aggregate be β%. After adjustment, the calculated values ​​are as follows:

[0109] Manufactured sand dosage: theoretical dosage × (1+α);

[0110] Coarse aggregate usage: theoretical usage × (1+β);

[0111] Water consumption: Theoretical water consumption - Sand moisture content - Stone moisture content.

[0112] Assuming the theoretical mix proportion (material usage per cubic meter of concrete) is cementitious material (C), water (W), manufactured sand (S), and coarse aggregate (G), in kg, the adjusted construction mix proportion is:

[0113] Manufactured sand dosage = S × (1 + α / 100);

[0114] Coarse aggregate usage = G × (1 + β / 100);

[0115] Water consumption = W - S × α / 100 - G × β / 100;

[0116] Cementitious material dosage = C (unchanged).

[0117] The tunnel excavation muck aggregate produced in a certain tunnel project was used. The design required the lining concrete to have a strength grade of C40, with no air content requirement. The aggregate used was granite, with a methylene blue (MB) value of 0.8, a stone powder content of 4.1%, no clay lump content, a crushing index of 17%, and a loose bulk density of 1490 kg / m³. 3 The fineness modulus is 2.7, the maximum particle size of the coarse aggregate is 31.5, and there are three types of gravel with different particle sizes: 5-10mm, 10-20mm, and 20-31.5mm. The gradation adopts the 442 gradation commonly used in engineering (particle size from largest to smallest).

[0118] C40 concrete was prepared in the on-site laboratory according to the design method proposed in Example 1, as shown in Table 6. The mix proportions were: water-cement ratio 0.48, cement 381.1 kg / m³. 3 The sand ratio is 45%, the granite powder content is 10%, and the coarse aggregate is 1076.4 kg / m³. 3 Water-reducing agent 5.4 kg / m³ 3The measured strength at 28 days was 45.1 MPa, and the drying shrinkage at 56 days was 301 × 10⁻⁶ MPa. -6 The drying shrinkage rate is relatively small and within the recommended range, meeting the actual requirements of the project.

[0119] Table 6. Mix proportions and performance testing of C40 concrete based on the design method of Example 1.

[0120]

[0121] The above embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the present invention, and the patent protection scope of the present invention should be defined by the claims.

Claims

1. A method for designing coordinated deformation of full-cavity slag concrete, characterized in that, The method includes: S1: Determine the strength grade and workability requirements of the concrete; wherein the workability requirements include the target slump value; S2: Cement, slag aggregate, stone powder, and water-reducing agent are used as raw materials. The cement is selected as silicate cement according to its strength grade. The slag aggregate includes manufactured sand and coarse aggregate. The manufactured sand is medium sand with a fineness modulus of 2.3-3.

0. The maximum particle size of the coarse aggregate is 31.5 mm. The stone powder is granite powder or limestone powder. The water-reducing agent is polycarboxylate-based and has a water reduction rate of 20%-30%. S3: Based on a pre-set database of raw materials, deformation properties, and mechanical properties, performance prediction is performed using a 56-day deformation prediction model and a 28-day mechanical prediction model. Combined with a genetic algorithm, multi-objective optimization is performed to output a recommended ratio range. The multi-objective optimization aims to maximize compressive strength and minimize shrinkage. S4: Conduct trial mixing according to the recommended mixing ratio range, and test the slump, 7-day compressive strength and 56-day drying shrinkage. If the slump does not meet the requirements, adjust the water-reducing agent dosage or water-cement ratio; if the 7-day compressive strength is less than 70% of the design value, readjust the mixing ratio; if the 56-day drying shrinkage is not within the recommended range, readjust the mixing ratio. S5: Detect the moisture content of manufactured sand and coarse aggregate, and adjust the amount of manufactured sand, coarse aggregate, and water according to the moisture content to obtain the construction mix proportion.

2. The method for coordinated deformation design of full-tunnel slag concrete as described in claim 1, characterized in that, In step S1, the strength grade of the concrete is C30-C60, and the target slump value is 100-220mm.

3. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S2, the content of the stone powder meets the following requirements: the content of non-reactive stone powder does not exceed 15%, and the content of reactive stone powder does not exceed 20%; the methylene blue value of the manufactured sand is ≤1.4, the crushing index is ≤20%, and the loose bulk density is ≥1400kg / m³. 3 .

4. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S3, the 56-day deformation prediction model is a neural network-based prediction model, and the 28-day mechanical prediction model is a gradient boosting tree-based prediction model. The recommended mix proportion range includes the following 28-day compressive strength range and 56-day shrinkage rate range for each strength grade: C30: 28-day compressive strength 30-40 MPa, 56-day shrinkage 351-480 × 10⁻⁶ MPa. -6 ; C40: 28-day compressive strength 40-50 MPa, 56-day shrinkage 304-455 × 10⁻⁶ MPa. -6 ; C50: 28-day compressive strength 50-60 MPa, 56-day shrinkage 287-439 × 10⁻⁶ MPa. -6 ; C60: 28-day compressive strength 60-70 MPa, 56-day shrinkage 250-368 × 10⁻⁶ MPa. -6 .

5. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S3, in the recommended mix ratio range output by the multi-objective optimization, the water-cement ratio, cement dosage, sand ratio, coarse aggregate dosage, and water-reducing agent dosage are fixed values, while the dosages of fly ash, mineral powder, and stone powder are adjustable ranges; the dosage range of stone powder is determined according to the lithology, as follows: C30: Granite powder 12.3-18.5%, limestone powder 9.5-14.8%; C40: Granite powder 8.3-14.4%, limestone powder 6.6-11.5%; C50: 6.3-10% granite powder, 5-8% limestone powder; C60: Granite powder 0-6.8%, Limestone powder 0-5.4%.

6. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, Step S3 also includes stone powder content correction, specifically including: using stone powder to replace fly ash and / or mineral powder by mass, and determining the replacement amount based on the replacement coefficient, which is calculated by the formula: in, The substitution coefficient, This represents the correlation coefficient between stone powder and fly ash or mineral powder. The standard deviation of stone powder content. This represents the standard deviation of fly ash or mineral powder content.

7. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S4, the adjustment method for the slump includes: When the slump is insufficient, increase the water-reducing agent dosage by 0.1% or fine-tune the water-cement ratio by ±0.01 each time; when the cohesiveness is poor, increase the sand ratio by 1% each time. When controlling the gas content, if the gas content does not meet the requirement of 3%-5%, the amount of gas-entraining agent should be adjusted by 0.005% each time.

8. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S4, cubic specimens are prepared for compressive strength testing, and prism specimens are prepared for shrinkage testing. The cubic specimens are cured for 7 days or 28 days, and the prism specimens are cured for 3 days and then moved into a drying oven at 20±2℃ and 60±5% humidity for curing.

9. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, In step S5, the adjustment formulas for the amount of manufactured sand, coarse aggregate, and water are as follows: Manufactured sand dosage = Theoretical manufactured sand dosage × (1 + α / 100); Coarse aggregate usage = Theoretical coarse aggregate usage × (1 + β / 100); Water consumption = Theoretical water consumption - Theoretical manufactured sand consumption × α / 100 - Theoretical coarse aggregate usage × β / 100; in, α The moisture content of the manufactured sand. β The moisture content of the coarse aggregate. α and β The numerical unit is percentage.

10. The method for coordinated deformation design of full-tunnel slag concrete according to claim 1, characterized in that, The tunnel slag aggregate is obtained by processing the tunnel slag produced during tunnel excavation. The coarse aggregate uses stones with three particle sizes: 5-10mm, 10-20mm, and 20-31.5mm.

Images