Alkali residue-free multi-source solid waste cementitious material proportioning optimization method and material

Abstract

The application belongs to the technical field of solid waste resource utilization, and discloses a method for optimizing the proportioning of a multi-source solid waste cementing material without alkali slag and the material. The method comprises the following steps: obtaining initial proportioning data of a plurality of solid waste raw materials and corresponding performance data, and determining the proportioning constraint range of each solid waste raw material according to the initial proportioning data and the performance data; introducing the initial proportioning data and the corresponding performance data as historical data into an augmented mixture design within the proportioning constraint range, generating supplementary proportioning points, and obtaining a modeling data set composed of the historical data points and the supplementary proportioning points; taking the proportioning of each solid waste raw material in the modeling data set as an independent variable, and taking the corresponding performance data as a response value, to establish a proportioning-performance response model; and based on the proportioning-performance response model, performing constraint optimization within the proportioning constraint range to obtain an optimized proportioning of the solid waste-based cementing material. The application can make full use of existing experimental data, reduce invalid repeated tests, and improve the proportioning optimization efficiency.

Description

Technical Field

[0001] This invention belongs to the field of solid waste resource utilization technology, specifically relating to a method and material for optimizing the proportion of alkali-free multi-source solid waste cementitious materials. Background Technology

[0002] With the development of industries such as mining, metallurgy, and coal-fired power plants, the generation of bulk industrial solid wastes, such as slag, steel slag, desulfurization gypsum, and fly ash, continues to increase. If these solid wastes are not effectively utilized, they will not only occupy a large amount of land resources but may also generate environmental problems such as dust, alkaline leachate, and potential heavy metal migration. Therefore, developing cementitious materials based primarily on multi-source industrial solid waste is an important direction for realizing the resource utilization of industrial solid waste and reducing the use of traditional cement.

[0003] Existing solid waste-based cementitious materials typically use slag, fly ash, steel slag, gypsum, red mud, alkaline slag, and magnesium slag as main raw materials, prepared through alkali activation, sulfate activation, or multi-component composite activation. Among these, slag possesses potential hydraulic activity, steel slag provides a certain alkaline environment and calcium source, desulfurized gypsum provides sulfate activation, and fly ash can serve as an auxiliary pozzolanic material in later stages of the reaction. The synergistic effect among various solid waste components can, to some extent, improve the hydration reaction process and mechanical properties of the cementitious materials.

[0004] However, existing solid waste-based cementitious materials still have shortcomings in mix design and engineering applications. On the one hand, the mix proportion determination of most solid waste-based cementitious materials still relies mainly on empirical trial and error, single-factor experiments, or ordinary orthogonal experiments, which makes it difficult to fully reflect the synergistic and mutually restrictive relationships among multiple components such as slag, steel slag, desulfurization gypsum, and fly ash. On the other hand, the sum of the proportions of each solid component in the cementitious material is fixed, belonging to a typical mixed system. The components are not independent of each other, but rather there is a proportional constraint relationship where one increases while the other decreases. Therefore, traditional orthogonal experiments or single-factor experiments cannot accurately reveal the comprehensive impact of changes in component proportions on material strength.

[0005] Furthermore, existing mixture optimization methods often generate all test points at once, failing to fully utilize existing experimental data and easily leading to numerous repeated experiments and insufficient optimization efficiency. Therefore, it is necessary to establish an optimization method that can incorporate existing experimental data, is applicable to multi-component ratio-constrained systems, and can obtain the optimal proportions of solid waste-based cementitious materials. Summary of the Invention

[0006] The purpose of this invention is to provide a method and material for optimizing the proportion of alkali-free multi-source solid waste cementitious materials, in order to solve the problems of existing solid waste-based cementitious material proportion design relying on experience and trial and error, requiring a large number of experiments, and making it difficult to reflect the proportional constraints and synergistic effects of multiple components.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: On the one hand, the present invention provides a method for optimizing the proportion of alkali-free multi-source solid waste cementitious materials.

[0008] S1. Raw material condition control and raw material parameter acquisition: Slag powder, steel slag powder, desulfurized gypsum and fly ash are pretreated to obtain solid raw materials that meet the requirements for preparing cementitious materials; and the raw material parameters of slag powder, steel slag powder, desulfurized gypsum and fly ash are obtained, which are used to determine the initial proportioning design parameters.

[0009] Preferably, the pretreatment includes drying, grinding, and sieving.

[0010] Preferably, the raw material parameters include chemical composition parameters, mineral composition parameters, and physical property parameters.

[0011] More preferably, the chemical composition parameters include at least CaO content and SiO2 content; the calcium-silicon ratio of the initial mixing system is calculated based on the CaO content and SiO2 content, and the calcium-silicon ratio is used as the basis for setting the initial mixing point.

[0012] S2, Initial proportions: Multiple initial mixing points are set according to the raw material parameters. Cementitious material specimens are prepared according to the component ratios corresponding to each initial mixing point, and the compressive strength data corresponding to each initial mixing point is obtained. Based on the initial mixing points and their compressive strength data, the mass percentage constraint range of slag powder, steel slag powder, desulfurized gypsum and fly ash is determined. Some initial mixing points and their corresponding compressive strength data are selected from the initial mixing points to form a historical dataset for subsequent augmented mixture design.

[0013] Preferably, the initial proportioning point is set using a single-factor proportioning method.

[0014] Preferably, the initial proportioning points include a first set of initial proportioning points and a second set of initial proportioning points; wherein, the first set of initial proportioning points is used to investigate the strength development law of cementitious materials under different calcium-silicon ratios by fixing the content of steel slag powder and desulfurized gypsum and changing the relative content of slag powder and fly ash; the second set of initial proportioning points is set on the basis of reducing the content of steel slag powder and desulfurized gypsum and constraining the content of slag powder and fly ash, and is used to determine the component constraint range of subsequent extended mixture design.

[0015] Preferably, based on the compressive strength development results at the initial mixing point, the mass percentage constraint ranges of slag powder, steel slag powder, desulfurized gypsum, and fly ash are determined as follows: 40%–70% slag powder, 25%–30% steel slag powder, 5%–10% desulfurized gypsum, and 0%–30% fly ash.

[0016] Preferably, historical data points are selected from the initial mix design points according to the following rules: First, the selected mix design points should be within the component constraints set by the subsequent augmented mix design, or be representative points near the boundary of the constraint space; second, the compressive strength corresponding to the selected mix design points should have stable test results that can be used for model fitting and can reflect the strength development trend at different ages; third, the selected mix design points should cover major mix design factors such as the relative content changes of slag powder and fly ash, and the content changes of steel slag powder and desulfurized gypsum. Initial mix design points that significantly deviate from the subsequent constraint range or have abnormal strength development in the later stages and are not representative will not be included as historical data points in the augmented mix design.

[0017] S3, Historical Data Augmented Mixing Design: Within the component mass percentage constraint, historical data points are introduced into the augmented mixing design model to generate new mixing design points, and the historical data points and new mixing design points are merged to form a mixing design dataset for response model construction.

[0018] Preferably: The newly added mixing design points include at least one of the following: vertex, center point, model point, duplicate item, and misfit check point.

[0019] Further preferred: There are 9 sets of historical data points and 12 sets of newly added mixing design points, which are combined to form 21 sets of mixing design points.

[0020] Within the component mass percentage constraint, historical data points are introduced into the augmented mixing design model to generate new mixing design points, and the historical data points and new mixing design points are merged to form a mixing design dataset for response model construction.

[0021] Specifically, the mass percentages of slag powder, steel slag powder, desulfurized gypsum, and fly ash are denoted as A, B, C, and D, respectively, satisfying A+B+C+D=1. Here, A ranges from 0.40 to 0.70, B from 0.25 to 0.30, C from 0.05 to 0.10, and D from 0 to 0.30. Within the above component constraints, previously selected historical data points are used as fixed design points. An augmented mixing design is employed to supplement and generate new mixing design points, ensuring that both historical data points and new mixing design points jointly cover the target mix proportion space.

[0022] The principle for generating new mixing design points is as follows: Under the premise of satisfying the mass percentage constraints of each component and the total constraint, priority is given to supplementing the boundary regions, central regions, and model validation regions of the constraint space, so that the modeling dataset can meet the needs of proportion-intensity response model fitting, repeatability error estimation, and fit-out testing. New mixing design points include one or more of the following: model points, repeatability points, and fit-out testing points. Model points are used to improve the fitting accuracy of the response model, repeatability points are used to estimate pure error, and fit-out testing points are used to determine whether the selected model form is suitable.

[0023] In a preferred embodiment, there are 9 sets of historical data points and 12 sets of newly added mixing design points, which are combined to form 21 sets of mixing design points.

[0024] The augmented mixture design described in this invention is based on existing historical experimental data. It expands the experimental space through mixture design theory, thereby improving the fitting ability and predictive reliability of the response model. Specifically, firstly, the constraint ranges for each component of the cementitious material are determined, namely, the mass fraction ranges of slag, steel slag, desulfurized gypsum, and fly ash are 0.40–0.70, 0.25–0.30, 0.05–0.10, and 0–0.30, respectively, while ensuring that the sum of the mass fractions of each component is 1. Subsequently, within the above-mentioned mixture constraint space, using nine sets of existing historical experimental data as the basic sample, the augmented mixture design method is used to generate new experimental points. These new experimental points cover the representative areas in the original design space, and model points, repetition points, and fit-out verification points are added.

[0025] S4. Response value acquisition: Cementitious material specimens were prepared according to the mass percentages of slag powder, steel slag powder, desulfurized gypsum, and fly ash corresponding to each mixing design point in the mixing design dataset. The compressive strength of each cementitious material specimen at a preset age was tested to obtain the strength response value corresponding to each mixing design point.

[0026] Preferably, the preset age range includes 3 days, 7 days, and 28 days.

[0027] Preferably, the preparation of the cementitious material specimen includes: weighing slag powder, steel slag powder, desulfurized gypsum and fly ash according to the corresponding mixing design point, mixing each solid raw material and stirring with water to obtain a slurry; pouring the slurry into a mold to form it, letting it stand, demolding it and curing it to the preset age.

[0028] Preferably, the water-to-binder ratio is 0.32.

[0029] Preferably, mechanical mixing is performed using a cement paste mixer. The mixing procedure is as follows: first, mix at low speed for 60 seconds to initially mix and fully wet the solid raw materials with water; then, mix at high speed for 30 seconds to enhance the uniformity of slurry dispersion; stop mixing for 60 seconds and scrape off the slurry from the mixing bowl wall and blades; finally, mix at high speed for another 60 seconds to obtain a uniform slurry. After the slurry is formed, it is allowed to stand at 20±2℃ for 24 hours before demolding; the demolded specimens are then placed under standard curing conditions for 3 days, 7 days, and 28 days. The preferred standard curing conditions are a temperature of 20±2℃ and a relative humidity of not less than 95%, or a standard curing chamber that can keep the specimens moist.

[0030] S5. Construction of the ratio-intensity response model: Using the mass fractions of slag powder, steel slag powder, desulfurized gypsum, and fly ash corresponding to each mixing design point in the mixing design dataset as independent variables, and the compressive strength corresponding to each mixing design point as the response value, a proportioning-strength response model for alkali-free slag multi-source solid waste cementitious materials is established.

[0031] To avoid the impact of differences in experimental batches, experimental times, or data sources between historical experimental data and subsequent augmented experimental data on the response model during the modeling process, this invention introduces block settings when performing augmented mixture design. Specifically, the nine sets of historical experimental data selected in the early stage are set as the first experimental block, namely Block 1; the 12 sets of newly generated mixture design points generated by the augmented mixture design are set as the second experimental block, namely Block 2. During modeling, block factors are included as auxiliary factors for non-mixture components in the model analysis to correct for possible systematic differences between different experimental batches; slag powder, steel slag powder, desulfurization gypsum, and fly ash are still included as mixture components in the proportion-intensity response model construction.

[0032] Block factors are used only to distinguish the data sources and test batches of historical data points from newly added mix design points; they are not involved in the calculation of cementitious material proportions, nor are they used as optimization target variables. If all historical data points and newly added mix design points are set in the same block, systematic differences between different batches may be incorrectly attributed to the residual term or mix component effects, leading to changes in response model coefficients, model significance, predicted values, and the final optimization results. Therefore, this invention preferably uses a Block1 and Block2 block approach to improve the stability and repeatability of the augmented mix design response model.

[0033] During model selection, priority was given to models with a p-value less than 0.05, a p-value greater than 0.05 for the lack-of-fit term, a low coefficient of variation, and a high signal-to-noise ratio. If multiple models met these criteria, the final model was determined by combining the coefficient of determination, the adjusted coefficient of determination, and the consistency plot between experimental and predicted values. Based on these rules, a linear mixture model was used for 3-day and 7-day compressive strength tests, while a third-order mixture model was used for 28-day compressive strength tests.

[0034] Preferably, 3-day compressive strength response models, 7-day compressive strength response models, and 28-day compressive strength response models are established respectively. The actual mass fractions of slag, steel slag, desulfurized gypsum, and fly ash are denoted as A, B, C, and D, respectively, where A+B+C+D=1, A represents slag, B represents steel slag, C represents desulfurized gypsum, and D represents fly ash.

[0035] Preferably, the 3d compressive strength response model and the 7d compressive strength response model adopt a linear mixing model.

[0036] Preferably, the 3d compressive strength response model and the 7d compressive strength response model are expressed as follows: in This represents the predicted 3D compressive strength. The values ​​represent the predicted 7-day compressive strength, all in MPa; A, B, C, and D represent the actual mass fractions of slag, steel slag, desulfurized gypsum, and fly ash, respectively.

[0037] Preferably, the 28-day compressive strength response model is a third-order mixture model fitted based on 21 sets of modeling data. During modeling, A, B, C, and D are the mixture component variables, and the 28-day compressive strength is the response value. Least squares regression fitting is used. The initial model form includes linear terms, binary interaction terms, and ternary interaction terms; subsequently, the model form is selected based on indicators such as model significance, lack-of-fit test, coefficient of determination, adjusted coefficient of determination, coefficient of variation, and signal-to-noise ratio. Finally, a third-order mixture model is determined to be used for the 28-day compressive strength.

[0038] Preferably, the 28-day compressive strength response model is expressed as follows: in This represents the predicted compressive strength at 28 days, in MPa. , , , , , Indicates a binary interaction term; , , , This represents a ternary interaction term.

[0039] S6. Response model selection and constraint optimization: The reliability of the proportion-strength response model was evaluated, and the response model that met the preset evaluation conditions was used for proportion optimization. Among them, within the mass percentage constraints of slag powder, steel slag powder, desulfurized gypsum and fly ash, the 3d compressive strength and 7d compressive strength predicted by the response model met the preset thresholds as constraints, and the maximization of 28d compressive strength was taken as the optimization objective. The optimal proportion of the alkali-free slag multi-source solid waste cementitious material was obtained by solving the problem.

[0040] Preferably, the reliability evaluation includes one or more of the following: model significance test, lack of fit test, coefficient of determination evaluation, and comparison of predicted and measured values.

[0041] Preferably, the P-values ​​of the 3d compressive strength model, the 7d compressive strength model, and the 28d compressive strength model are all less than 0.05, and the P-values ​​of the lack-of-fit terms are all greater than 0.05.

[0042] Preferably, 3d compressive strength and 7d compressive strength are used as auxiliary constraint indicators, and 28d compressive strength is used as the main optimization indicator.

[0043] Preferably, the constraint optimization is performed using a numerical optimization method. The actual mass fractions of slag, steel slag, desulfurized gypsum, and fly ash are denoted as A, B, C, and D, respectively, where A represents slag, B represents steel slag, C represents desulfurized gypsum, and D represents fly ash, and the following conditions are met: Within the component constraints described in this invention, each component satisfies the following conditions: A, B, C, and D are all expressed as mass fractions.

[0044] Under the above component constraints, with the 3d compressive strength and 7d compressive strength meeting preset ranges as auxiliary constraints, the primary optimization objective is to maximize the 28d compressive strength. The optimization objective function is expressed as: The constraints are: in, , and These represent the 3d, 7d, and 28d compressive strengths predicted by the response model, respectively, in MPa.

[0045] S7, Optimal Proportion Output: Output the mass percentage combination of slag powder, steel slag powder, desulfurized gypsum and fly ash that satisfies the constraints and achieves the optimized 28d compressive strength target, as the preferred ratio for alkali-free slag multi-source solid waste cementitious materials.

[0046] Preferably, the preferred proportions are: 61.83% slag powder, 28.17% steel slag powder, 10% desulfurized gypsum, and 0% fly ash.

[0047] Preferably, the predicted 3-day compressive strength corresponding to the preferred ratio is 17.0003 MPa, the predicted 7-day compressive strength is 23.3766 MPa, and the predicted 28-day compressive strength is 30.1715 MPa.

[0048] Specifically, the optimized mix ratio is substituted into the above compressive strength response model to calculate its predicted compressive strength values ​​at 3d, 7d, and 28d. In the optimized mix ratio, the actual mass fractions of slag, steel slag, desulfurized gypsum, and fly ash are respectively: In this context, A represents slag, B represents steel slag, C represents desulfurized gypsum, and D represents fly ash.

[0049] Taking the calculation of the predicted 28-day compressive strength as an example, the above values ​​are substituted into the 28-day compressive strength response model: Substitution , , , Then, we can obtain: S8, Optional preferred ratio verification In a preferred embodiment, a cementitious material specimen is prepared according to a preferred ratio and its compressive strength at a preset age is tested to verify the predictive reliability of the response model.

[0050] Preferably, the experimental values ​​of the 3-day compressive strength corresponding to the preferred ratio are 14.43 MPa, 25.93 MPa, and 32.34 MPa; the corresponding relative errors are 15.12%, 10.92%, and 7.19%, respectively.

[0051] Preferably, based on the verification results, the 28-day compressive strength is used as the main optimization basis, and the 3-day compressive strength and 7-day compressive strength are used as auxiliary constraint indicators.

[0052] On the other hand, the present invention provides an alkali-free slag multi-source solid waste cementitious material, the proportions of which are obtained by the aforementioned method. Specifically, the alkali-free slag multi-source solid waste cementitious material is prepared by mixing slag powder, steel slag powder, desulfurized gypsum and optional fly ash with water. The mass percentage of solid components in the alkali-free slag multi-source solid waste cementitious material is as follows: slag powder 61.83%, steel slag powder 28.17%, desulfurized gypsum 10%, and fly ash 0%.

[0053] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses slag powder, steel slag powder, desulfurized gypsum and fly ash to form an alkali-free multi-source solid waste cementitious material system. Without introducing additional alkali slag or magnesium slag, it can utilize the alkaline components and calcium source provided by steel slag and the sulfate activation effect provided by desulfurized gypsum to synergistically promote the hydration reaction of slag and improve the resource utilization value of multi-source solid waste.

[0054] 2. This invention transforms the initial proportion-strength data into a historical dataset and component constraint space, and introduces the historical dataset into the augmented mixture design process, which can make full use of existing experimental data, reduce invalid repeated experiments, and improve the efficiency of proportion optimization.

[0055] 3. This invention targets a mixing system in which the sum of the mass percentages of slag powder, steel slag powder, desulfurized gypsum, and fly ash is fixed. Through augmented mixing design and response model construction, it can reflect the proportional constraints between the components, as well as the synergistic and mutually restrictive relationships between multiple components.

[0056] 4. This invention establishes compressive strength response models for different ages, and uses 3d compressive strength and 7d compressive strength as auxiliary constraint indicators, and 28d compressive strength as the main optimization indicator. It can screen the optimal ratio that takes into account both early strength development and later strength performance within a limited ratio range.

[0057] 5. The preferred formulation obtained by this invention exhibits good compressive strength at 28 days. In the verification experiment, the experimental value of the 28-day compressive strength of the preferred formulation was 32.34 MPa, with a relative error of 7.19%, indicating that the established model has good reliability for predicting the strength in the later stage. Attached Figure Description

[0058] Figure 1 The graph shows the compressive strength test results of the first round of single-factor experimental formulations; Figure 2 The graph shows the compressive strength test results of the second round of single-factor experimental formulations; Figure 3 A comparison chart of the experimental and predicted values ​​of 3d compressive strength at 21 test sites; Figure 4 This is a 3D compressive strength response tracking graph; Figure 5 A comparison chart of the experimental and predicted values ​​of the 7-day compressive strength at 21 test sites; Figure 6 This is a 7-day compressive strength response tracking graph; Figure 7 A comparison chart of the experimental and predicted values ​​of the 28-day compressive strength at 21 test sites; Figure 8 This is a 28-day compressive strength response tracking graph; Figure 9 The cross section and optimal point plot of the multi-response superposition feasible domain when the fly ash mass fraction D=0 are shown. Detailed Implementation

[0059] To enable those skilled in the art to better understand the present application, the technical solutions in specific embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by those skilled in the art.

[0060] Example 1 This embodiment illustrates the preparation of specimens of alkali-free multi-source solid waste cementitious materials and the method for testing their compressive strength.

[0061] The solid raw materials used in this embodiment include slag powder, steel slag powder, desulfurized gypsum, and fly ash. Each raw material is dried, ground, and sieved before use to reduce the impact of moisture content and particle agglomeration on batching accuracy and test results. The chemical composition, mineral composition, and physical properties of each processed raw material are analyzed. The chemical composition parameters include at least CaO content and SiO2 content, and the physical property parameters include particle size or specific surface area parameters that characterize the particle state of the raw material.

[0062] The total mass percentage of slag powder, steel slag powder, desulfurized gypsum, and fly ash in the cementitious material is 100%, and the water-cement ratio is set at 0.32. Slag powder, steel slag powder, desulfurized gypsum, and fly ash are weighed according to the proportions corresponding to each test point. The weighed solid raw materials are placed into sample bags and shaken for 30 seconds to allow the solid raw materials to be initially mixed evenly.

[0063] The pre-mixed solid raw materials and the corresponding mass of water are added to a cement paste mixer for mechanical mixing. The specific mixing procedure is as follows: first, mix at low speed for 60 seconds to pre-mix and fully wet the solid raw materials with water; then mix at high speed for 30 seconds to improve the dispersion uniformity of the slurry; stop mixing for 60 seconds, during which time the slurry adhering to the mixing bowl wall and mixing blades is scraped off; finally, mix at high speed for 60 seconds to obtain a uniform slurry.

[0064] Pour the mixed slurry into a mold measuring 30mm × 30mm × 50mm, and compact and level the slurry as necessary. After molding, cover the mold with plastic wrap to prevent excessive evaporation of moisture from the specimen surface. After standing at 20℃ for 24 hours, demold the specimen. Place the demolded specimen under standard curing conditions of 20℃ and relative humidity not less than 95%, or in a standard curing chamber that keeps the specimen moist, and cure for 3 days, 7 days, and 28 days, respectively.

[0065] The compressive strength test was conducted according to the basic test principles of GB / T 17671-2021 "Test Method for Strength of Cement Mortar (ISO Method)" and in conjunction with the specimen dimensions in this embodiment. Before testing, the specimens cured to the specified age were removed, and any adhering water on the surface was wiped off. The specimens were then inspected for defects such as obvious missing corners, cracks, honeycombing, or uneven bearing surfaces. The specimen meeting the test requirements was placed at the center of the pressure plate of the compression testing machine, ensuring that the pressure surface of the specimen was perpendicular to the loading direction and that the upper and lower pressure surfaces of the specimen were in full contact with the pressure plate of the compression testing machine. After starting the compression testing machine, a continuous and uniform load was applied until the specimen failed, and the maximum failure load F was recorded.

[0066] The compressive strength of the specimen is calculated according to the following formula: The value of F represents the compressive strength of the specimen, in MPa; F represents the maximum failure load of the specimen, in N; and A represents the actual compressive area of ​​the specimen, in mm². 2 For the same mix proportion and age, the compressive strength is taken as the arithmetic mean of the test results of the effective specimens, which is used as the compressive strength response value corresponding to that mix proportion point.

[0067] Unless otherwise stated, the cementitious material specimens in subsequent embodiments were prepared and tested in accordance with the water-cement ratio, stirring procedure, molding method, curing conditions and compressive strength testing method described in this embodiment.

[0068] Example 2 This embodiment uses two rounds of single-factor proportioning tests to determine the initial proportioning variation law of slag powder, steel slag powder, desulfurized gypsum and fly ash, the component constraint range of subsequent mixing design, and historical data points for augmented mixing design.

[0069] First, based on the chemical composition data of each raw material obtained in Example 1, and combined with the mass fractions of slag powder, steel slag powder, desulfurized gypsum, and fly ash in different proportions, the calcium-silicon ratio of each initial cementitious material system was calculated. The same calculation caliber was used for each proportion, with the CaO and SiO2 content contributed by each raw material in the system as the calculation basis. The calcium-silicon ratio is used to characterize the relative relationship between calcium-containing and silicon-containing components in the proportioning system and serves as a reference parameter for setting the first round of single-factor proportioning points.

[0070] In the first round of single-factor experiments, the steel slag powder content was fixed at 30% and the desulfurized gypsum content at 10%. By changing the relative content of slag powder and fly ash, cementitious material systems with different calcium-silicon ratios were formed to investigate the development law of the 3-day, 7-day, and 28-day compressive strength of cementitious materials under different calcium-silicon ratios. The mix proportions of the first round of single-factor experiments are shown in Table 1.

[0071] Table 1: Proportions for the first round of single-factor experiments

[0072] The chemical composition of each raw material is shown in Table 2.

[0073] Table 2: Chemical composition of raw materials

[0074] Calculate the total mass content of CaO and SiO2 in the mixed cementitious system based on the actual blending ratio of different raw materials: Based on the actual blending ratios of different raw materials, the total mass content of CaO and SiO2 in the mixed cementitious system was calculated respectively: Let be the mass fraction of the i-th raw material in the cementitious material. and These represent the mass percentages of CaO and SiO2 in the i-th raw material, respectively.

[0075] Subsequently, the molar calcium-to-silicon ratio of the mixed system was calculated based on the molar mass conversion of CaO and SiO2: Cementitious material specimens were prepared according to the proportions listed in Table 1, and their compressive strength at 3d, 7d, and 28d was tested. The results are as follows: Figure 1 As shown.

[0076] The results of the first round of single-factor experiments showed that as the slag powder content decreased and the fly ash content increased, the calcium-silicon ratio and compressive strength development of the cementitious materials changed significantly. Some low-slag, high-fly ash mixes exhibited slow compressive strength growth or abnormal strength development in the later stages after 28 days. This indicates that under the conditions of 30% steel slag powder and 10% desulfurized gypsum content, further reducing the slag powder content and increasing the fly ash content may lead to an imbalance in the proportions of active components, alkaline components, and sulfate components in the system, which is detrimental to the continued progress of the later hydration reaction.

[0077] Based on the results of the first round of single-factor experiments, the second round of single-factor experiments reduced the content of steel slag powder and desulfurization gypsum, fixing the steel slag powder at 25% and the desulfurization gypsum at 5%, while limiting the content of slag powder to 40%–70% and the content of fly ash to 0%–30%. By continuously changing the relative content of slag powder and fly ash, the compressive strength development of each mix ratio was further investigated. The mix ratios for the second round of single-factor experiments are shown in Table 3.

[0078] The sample combinations for the second round of experiments are shown in the table below: Table 3: Proportions for the Second Round of Single-Factor Experiments

[0079] Cementitious material specimens were prepared according to the proportions listed in Table 3, and their compressive strength was tested at 3d, 7d, and 28d. The results are as follows: Figure 2 As shown.

[0080] The results of the second round of single-factor tests showed that the 28-day compressive strength of the cementitious materials in groups 5 to 11 all developed normally based on their early strength. Among them, when the mass percentages of slag powder, steel slag powder, desulfurized gypsum, and fly ash were 70%, 25%, 5%, and 0%, respectively, the 28-day compressive strength reached 29.875 MPa, which was the optimal mix ratio in the single-factor test stage.

[0081] Based on the results of two rounds of single-factor experiments, the mass percentage ranges of slag powder, steel slag powder, desulfurization gypsum, and fly ash in the subsequent extended mixture design were determined to be 40%–70%, 25%–30%, 5%–10%, and 0%–30%, respectively, with the sum of the mass percentages of the four solid components being 100%. Nine sets of experimental data (numbered 1 to 11) were selected as historical data points from the initial experimental points numbered 1–11. The selection criteria for the historical data points are as follows: First, numbers 1, 2 and numbers 5 to 11 are all within the component constraints set by the subsequent extended mixing design, or at the boundary of the component constraint space, and can be used to characterize the component changes within the target proportioning space.

[0082] Second, the ratio changes of slag powder from 70% to 40% and fly ash from 0% to 30% in sections 5 to 11 can reflect the influence of the relative content of slag powder and fly ash on the compressive strength at different ages.

[0083] Third, the proportions of steel slag powder (30%) and desulfurized gypsum (10%) corresponding to No. 1 and No. 2 can be used to supplement the characterization of the proportions of steel slag powder and desulfurized gypsum near their upper limits.

[0084] Fourth, the above nine sets of test data have compressive strength test results that can be used for model fitting and can reflect the strength development trend at different ages.

[0085] The slag powder content of No. 3 and No. 4 is 30% and 15% respectively, which is lower than the 40% lower limit specified in the subsequent extended mixing design; the fly ash content of No. 4 is 45%, which is higher than the 30% upper limit of the subsequent constraint range, and the later strength representativeness of the corresponding proportions is weak. Therefore, No. 3 and No. 4 will not be introduced into the subsequent extended mixing design as historical data points.

[0086] Thus, the initial screening of the target proportion space was completed through two rounds of single-factor experiments, and a historical dataset consisting of 9 sets of existing experimental data was formed, providing a basis for the historical data augmentation mixing design of Example 3.

[0087] Example 3 Based on the nine sets of historical data formed in Example 2, this embodiment performs an augmented mixing design and establishes a proportion-strength response model based on the compressive strength data of the historical data points and the newly added mixing design points.

[0088] Let the actual mass fractions of slag powder, steel slag powder, desulfurized gypsum, and fly ash be denoted as A, B, C, and D, respectively, where A, B, C, and D satisfy the following: The constraint ranges for each component are as follows: Nine sets of experimental data corresponding to numbers 1, 2, 5, 6, 7, 8, 9, 10, and 11 selected in Example 2 were introduced into the augmented mixture design as fixed historical data points. The addition of new mixture design points was based on the principle of supplementing the constraint space boundary region, central region, and model validation region that were not fully covered by historical data, so that the combination of historical data points and new mixture design points could meet the requirements of response model fitting, pure error estimation, and fit test.

[0089] This embodiment generates 12 new mixing design points, including 10 model points, 1 set of repeat points, and 1 set of fit check points. The model points are used to improve the fitting and predictive ability of the response model within the target mix space; the repeat points are used to estimate the pure error by combining the repeated test results of the same mix proportion points; and the fit check points are used to evaluate whether there is a significant fit between the selected response model form and the actual mix proportion-intensity relationship.

[0090] After merging 9 sets of historical data points with 12 sets of newly added mixing design points, a total of 21 sets of modeling test points were formed. The structural type, spatial type and group allocation of each test point are shown in Table 4.

[0091] Table 4: Experimental ratios for 21 test sites

[0092] Test points 1 to 9 in Table 4 are from the initial tests already completed in Example 2, and are set as the first test group Block 1; Test points 10 to 21 are tests supplemented to this example based on the extended compound design, and are set as the second test group Block 2.

[0093] The purpose of setting up blocks is to distinguish the data sources and batches of historical and newly added experiments, and to correct for systematic differences caused by experimental batches, experimental time, or other non-proportioning factors during model analysis. Block factors participate in statistical analysis as auxiliary factors for non-mixed components. They are not part of the solid components of cementitious materials, do not participate in the sum constraint of A+B+C+D=1, and are not used as target variables for subsequent proportioning optimization.

[0094] As shown in Table 4, the newly added test points include the vertices of the constraint space, the center point of the plane, the top center point, and the axial verification point. Test point 15 uses the same mix ratio as test point 11 and can be used for repeatability error evaluation; the newly added test point 16 uses the same mix ratio as the historical test point 9 and can also be used to evaluate the response differences between different test batches. Test point 17 is located inside the target mix ratio space and serves as an axial misfit verification point, used to determine whether the selected model can reasonably describe the relationship between mix ratio changes and compressive strength.

[0095] Since Table 4 has fully disclosed the specific component ratios of each newly added mixing design point, those skilled in the art can directly repeat this embodiment according to the 12 sets of newly added proportions listed in Table 4, without relying on a specific software version to automatically generate the exact same test points.

[0096] According to the proportions listed in Table 4 for each test point, cementitious material specimens were prepared using the method described in Example 1, and the compressive strength at 3d, 7d, and 28d for each test point was tested. The actual mass fractions of slag powder, steel slag powder, desulfurized gypsum, and fly ash at the 21 test points were used as mixture component variables, and the compressive strength at the corresponding age was used as the response value for regression fitting.

[0097] During the modeling process, test points 1–9 were designated as Block 1, and test points 10–21 were designated as Block 2. Block factors were incorporated into the model analysis as non-mixing auxiliary factors to correct for potential systematic differences between historical and newly added test batches. After block correction, a response model for proportion prediction and optimization was formed using mixture components A, B, C, and D.

[0098] During model selection, a comprehensive evaluation is conducted, considering model significance, significance of the lack-of-fit term, coefficient of determination, adjusted coefficient of determination, coefficient of variation, signal-to-noise ratio (SNR), and the consistency between predicted and experimental values. Models with a p-value less than 0.05, a p-value greater than 0.05 for the lack-of-fit term, a relatively low coefficient of variation, and a relatively high SNR are prioritized. When multiple candidate models simultaneously meet these criteria, the final model form is determined by further considering the coefficient of determination, adjusted coefficient of determination, and the consistency plot between predicted and experimental values.

[0099] According to the above screening rules, the 3d compressive strength and 7d compressive strength adopt the linear mixing model, while the 28d compressive strength adopts the third-order mixing model that includes linear terms, binary interaction terms, and ternary interaction terms.

[0100] The 3D compressive strength response model is as follows: The 7d compressive strength response model is as follows: The 28-day compressive strength response model is as follows: The block designation is primarily used to eliminate potential systematic errors between different batches of experimental data and improve the accuracy of the response model fitting. Its role is as an error control factor in the model building process, rather than a variable related to the composition of the cementitious materials. Therefore, when performing response model fitting and analysis of variance in Design-Expert software, the block designation is already involved in the model correction process. However, in the final response equation expressed in terms of actual mix components, only the proportions of slag (A), steel slag (B), desulfurized gypsum (C), and fly ash (D) are retained; the block designation is not listed separately.

[0101] The 3-day, 7-day, and 28-day compressive strength response models disclosed in this application are all prediction models established based on actual component variables, used to predict corresponding compressive strength values ​​according to different cementitious material ratios. The block design does not change the ratio relationship between cementitious material components and therefore is not used as a variable in the optimization solution process.

[0102] , representing the predicted compressive strength values ​​at 3d, 7d, and 28d respectively, all in MPa; A, B, C, and D represent the actual mass fractions of slag powder, steel slag powder, desulfurized gypsum, and fly ash respectively; AB, AC, AD, BC, BD, and CD represent the binary interaction terms of the corresponding components; ABC, ABD, ACD, and BCD represent the ternary interaction terms of the corresponding components.

[0103] The model coefficients described above were obtained by fitting 21 sets of modeling data using the least squares method. The modeling and subsequent numerical calculations used unrounded internal regression coefficients. The regression coefficients listed in the manual are rounded to five decimal places. Therefore, manual calculations using the publicly available coefficients may result in discrepancies in the last decimal place compared to the output of the modeling software.

[0104] The results of the model variance analysis and statistical evaluation are shown in Table 5.

[0105] Table 5: Analysis of Variance and Statistical Information of the Fitting Model

[0106] Table 5 shows that the P-values ​​of the compressive strength response models for 3d, 7d, and 28d are 0.0016, 0.0005, and 0.0186, respectively, all less than 0.05, indicating that the selected model terms significantly explain the overall compressive strength at the corresponding ages. The P-values ​​of the models' lack-of-fit terms are 0.4233, 0.7786, and 0.4021, respectively, all greater than 0.05, indicating that no significant lack of fit was detected at the current level of repeatability error, and the selected model form can be used to describe the overall law of compressive strength variation with mix proportion at the corresponding ages.

[0107] The coefficient of determination (COD) for the 3-day compressive strength model was 0.6045, the adjusted COD was 0.5303, the coefficient of variation was 13.97%, and the signal-to-noise ratio (SNR) was 10.077. The COD for the 7-day compressive strength model was 0.6608, the adjusted COD was 0.5972, the COD was 13.04%, and the SNR was 11.767. These results indicate that the 3-day and 7-day compressive strengths are significantly affected by the early activity of the raw materials, the initiation state of the hydration reaction, the uniformity of mixing, and fluctuations in molding and curing. The established linear model is more suitable for characterizing the overall trend of early and mid-term strength changes and can serve as an auxiliary constraint model for subsequent proportioning optimization.

[0108] The coefficient of determination of the 28-day compressive strength model is 0.9680, the adjusted coefficient of determination is 0.8987, the coefficient of variation is 5.72%, and the signal-to-noise ratio is 13.750. This indicates that the 28-day compressive strength model has a high interpretability of the modeling data, and the predicted signal is relatively sufficient relative to the experimental error, making it suitable as the main response model for subsequent mix optimization.

[0109] like Figure 3 and Figure 5 As shown, the predicted values ​​of 3d and 7d compressive strength generally vary with the experimental values, but there are deviations between some experimental points and the predictive-experimental value agreement line, indicating that the prediction accuracy of the early and mid-term strength models is relatively limited. Figure 7 As shown, the predicted values ​​of 28-day compressive strength are relatively consistent with the experimental values, which corresponds to the high coefficient of determination and low coefficient of variation of the 28-day model.

[0110] Depend on Figure 4 , Figure 6 and Figure 8 The response tracking results show that, within the component range defined in this embodiment, increasing the relative content of slag powder is generally beneficial to the development of 28-day compressive strength. When the fly ash content is high, its contribution to the active reaction is insufficient to fully compensate for the impact caused by the reduction of slag powder, and may exhibit a certain dilution effect. Steel slag powder and desulfurized gypsum can provide alkaline components, calcium sources, and sulfate activation effects, respectively, within an appropriate content range. However, when their relative content is too high, it may affect the stable development of the later hydration reaction. The above variation patterns are basically consistent with the single-factor test results of Example 2.

[0111] The complete component data of 21 test sites and the experimental values ​​of compressive strength at 3d, 7d and 28d are shown in Table 6.

[0112] Table 6: Compressive strength data of 21 test points

[0113] The 21 sets of data listed in Table 6 constitute the complete fitting dataset for the above response model. Among them, test points 1 to 9 are completed historical data points, and test points 10 to 21 are newly implemented data points after augmentation design. Test points 11 and 15 have the same ratio, as do historical test point 9 and newly added test point 16. The corresponding repeated test results can be used to evaluate the pure error and the fluctuation between different test batches.

[0114] This embodiment utilizes 9 existing historical experimental data sets and adds only 12 new experimental sets, resulting in 21 sets of modeling data that meet the requirements for model fitting, repeatability error estimation, and failure-to-fit test. Compared to re-implementing all 21 sets of experiments, this embodiment reduces the number of new experiments by 9, representing a reduction of: This demonstrates that incorporating existing experimental data as fixed historical data points into augmented mixture design can reduce invalid repeated experiments while meeting the requirements for establishing response models.

[0115] Example 4 This embodiment optimizes the mix proportions based on the 3d, 7d, and 28d compressive strength response models established in Embodiment 3, and experimentally verifies the optimized mix proportions.

[0116] The actual mass fractions of slag powder, steel slag powder, desulfurization gypsum, and fly ash are denoted as A, B, C, and D, respectively. The proportion optimization satisfies the following component constraints: Using the 3d and 7d compressive strengths meeting a preset response range as auxiliary constraints, and maximizing the 28d compressive strength as the primary optimization objective, the optimization problem is expressed as: Simultaneously satisfy: Among them, 20.2583 MPa and 25.3458 MPa are the higher measured values ​​of 3-day and 7-day compressive strength in the 21 sets of modeling data, respectively. Setting the above response range allows numerical optimization to be carried out within the response range covered by the modeling data, avoiding over-extrapolation of the model to response regions lacking experimental data support.

[0117] Numerical optimization methods were used to search for a mix design that satisfies the auxiliary constraints of 3d and 7d compressive strength and maximizes the predicted value of 28d compressive strength within the aforementioned component constraint space. The optimization settings for each component variable and response objective are shown in Table 7.

[0118] Table 7: Optimization Setting Conditions for Alkali-Free Multi-Source Solid Waste Cementitious Materials

[0119] like Figure 9 As shown, by superimposing the 3d compressive strength and 7d compressive strength constraints onto the mixing design space, a feasible mix design region that simultaneously meets the early and mid-term strength requirements is formed within the component constraint range. Figure 9 The feasible region in the model indicates that the model predicts a 3-day compressive strength of not less than 17 MPa and a 7-day compressive strength of not less than 22 MPa for the corresponding mix proportions; mix proportions located outside the feasible region have at least one auxiliary constraint that cannot be met.

[0120] Within the feasible area, numerical optimization was performed with the objective of maximizing the predicted 28-day compressive strength, resulting in the optimal mix ratio: 61.83% slag powder, 28.17% steel slag powder, 10.00% desulfurized gypsum, and 0% fly ash. Expressed as mass fractions: A=0.6183, B=0.2817, C=0.1000, D=0 Substituting the preferred ratio into the response model established in Example 3, the predicted 3-day compressive strength was 17.000 MPa, the predicted 7-day compressive strength was 23.376 MPa, and the predicted 28-day compressive strength was 30.173 MPa.

[0121] It should be noted that the numerical optimization and response value calculation use the unrounded model coefficients and unrounded matching variables saved during the modeling process. The model coefficients and optimal matching ratios disclosed in the manual have all been processed to retain the number of digits. Therefore, when using the integer values ​​listed in the manual for manual calculation, the results may have slight differences in the last digit compared with the predicted values ​​output by the software.

[0122] The optimal fly ash mass fraction in the preferred ratio is 0, indicating that under the raw material conditions, component range, and strength constraints set in this embodiment, the optimal solution predicted by the model is located at the lower limit boundary of fly ash content. This result does not mean that all solid waste cementitious material systems do not require the addition of fly ash, but rather it is the optimization result obtained for the raw materials and constraints of this embodiment.

[0123] Further experimental verification of the optimal ratio was conducted. The results are shown in Table 8. Table 8: Experimental Validation Results of the Model

[0124] As shown in Table 8, the predicted 3-day compressive strength of the optimized mix ratio is 17.0003 MPa, the experimental value is 14.43 MPa, and the relative error is 15.12%; the predicted 7-day compressive strength is 23.3766 MPa, the experimental value is 25.93 MPa, and the relative error is 10.92%; the predicted 28-day compressive strength is 30.1715 MPa, the experimental value is 32.34 MPa, and the relative error is 7.19%.

[0125] The experimental 28-day compressive strength value was higher than the model prediction value, with a relative error of 7.19%, indicating that the 28-day compressive strength model has good predictive ability for the later strength of the optimized mix ratio. The experimental 7-day compressive strength value was also higher than the model prediction value, indicating that the optimized mix ratio has good mid-term strength development ability.

[0126] The experimental 3-day compressive strength value was lower than the model prediction value and did not reach the 17 MPa prediction constraint set by the optimized model. This indicates that the 3-day compressive strength is greatly affected by factors such as the early activity of the raw materials, the initiation rate of the hydration reaction, the molding state of the specimen, and the curing conditions. The actual prediction accuracy of the 3-day linear model is relatively limited. Therefore, the 17 MPa in this embodiment is an auxiliary constraint set for the model prediction value during the numerical optimization process and should not be interpreted as meaning that the measured 3-day compressive strength of the material prepared using the optimized ratio will necessarily be no less than 17 MPa under all conditions.

[0127] Based on the comprehensive model statistical evaluation and verification test results, this invention uses the 28-day compressive strength model, which has relatively high predictive stability, as the main optimization basis, and the 3-day and 7-day compressive strength models as auxiliary constraint tools for screening feasible areas of the mix proportion. The optimal mix proportion obtained by the above method is 61.83% slag powder, 28.17% steel slag powder, 10.00% desulfurized gypsum, and 0% fly ash. Its verification test showed a 28-day compressive strength of 32.34 MPa.

[0128] Based on the combined results of single-factor experiments, augmented mixture design, response model analysis, and verification experiments, this invention can effectively screen the optimal formulation of cementitious materials for multi-source solid waste without alkali residue with fewer experiments. Compared with traditional empirical trial-and-error or single-factor optimization methods, this invention not only makes full use of existing experimental data but also reveals the synergistic effects and proportional constraints among multiple components. The optimal formulation obtained by this optimization method exhibits good compressive strength at 28 days and has high value for solid waste resource utilization.

[0129] Finally, it should be noted that the described embodiments are merely some, not all, of the embodiments of the present invention. Those skilled in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the claims and their equivalents; that is, all other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

Claims

1. A method for optimizing the proportioning of alkali-free, multi-source solid waste cementitious materials, characterized in that, Includes the following steps: Initial proportioning data and corresponding performance data of various solid waste raw materials were obtained, and the proportioning constraints of each solid waste raw material were determined based on the initial proportioning data and performance data. The initial proportioning data was set through single-factor proportioning experiments, which included: a first group of initial proportioning experiments, which investigated the compressive strength development law of cementitious materials under different calcium-silicon ratios by changing the relative amounts of slag powder and fly ash while fixing the amounts of steel slag powder and desulfurized gypsum; and a second group of initial proportioning experiments, which were set up by reducing the amounts of steel slag powder and desulfurized gypsum while constraining the amounts of slag powder and fly ash, to determine the component constraint range for subsequent extended mixture design. Within the constraints of the mixing ratio, the initial mixing ratio data and its corresponding performance data are introduced into the augmented mixing design as historical data to generate supplementary mixing ratio points and obtain a modeling dataset composed of historical data points and supplementary mixing ratio points. Among them, the supplementary mixing ratio points include one or more of the following: vertex, center point, model point, repetition point, and misfit verification point. The modeling dataset includes 9 sets of historical data points and 12 sets of supplementary mixing ratio points, forming a total of 21 sets of mixing ratio points. Using the proportions of each solid waste material in the modeling dataset as independent variables and the corresponding performance data as response values, a proportion-performance response model is established. Based on the ratio-performance response model, constraint optimization is performed within the ratio constraint range to obtain the optimal ratio of solid waste-based cementitious materials.

2. The optimization method according to claim 1, characterized in that, Before obtaining the initial proportioning data, various solid waste raw materials were dried, ground, and screened, and their chemical composition parameters, mineral composition parameters, and physical property parameters were obtained. Among them, the chemical composition parameters included CaO content and SiO2 content, and the initial proportioning data were determined based on the CaO content and SiO2 content.

3. The optimization method according to claim 1, characterized in that, The proportioning range is as follows: 40%–70% slag powder, 25%–30% steel slag powder, 5%–10% desulfurized gypsum, and 0%–30% fly ash, and the sum of the mass percentages of slag powder, steel slag powder, desulfurized gypsum, and fly ash is 100%.

4. The optimization method according to claim 1, characterized in that, Historical data are selected from the initial proportioning data and its corresponding performance data based on one or more of the following conditions: The mixing point is within the mixing constraint range; The proportioning points cover the boundary area of ​​the proportioning constraint range; The performance data corresponding to the mixing ratio point shows a normal development trend; The mixing point can represent the relative dosage changes between different solid waste raw materials.

5. The optimization method according to claim 1, characterized in that, Performance data includes compressive strength data, which includes 3d compressive strength, 7d compressive strength and 28d compressive strength; The ratio-performance response model includes a 3d compressive strength response model, a 7d compressive strength response model, and a 28d compressive strength response model. The 3d and 7d compressive strength response models are linear models, while the 28d compressive strength response model is a third-order model.

6. The optimization method according to claim 1, characterized in that, It also includes reliability evaluation of the ratio-performance response model, which includes one or more of the following: model significance test, lack of fit test, coefficient of determination evaluation, adjusted coefficient of determination evaluation, coefficient of variation evaluation, signal-to-noise ratio evaluation, and comparison of predicted and measured values.

7. The optimization method according to claim 6, characterized in that, When the proportion-performance response model meets the preset reliability evaluation conditions, it is used for constraint optimization. The constraint optimization includes: using the 3d compressive strength and 7d compressive strength predicted by the response model to meet the preset threshold as auxiliary constraints, and taking the maximization of 28d compressive strength as the optimization objective, to output the optimal proportion of solid waste-based cementitious materials.

8. A non-alkali residue multi-source solid waste cementitious material, characterized in that, The alkali-free slag multi-source solid waste cementitious material is prepared by using the optimization method described in any one of claims 1-7 to determine the proportions, and is made by mixing slag powder, steel slag powder, desulfurized gypsum and optional fly ash with water; the mass percentage of solid components in the alkali-free slag multi-source solid waste cementitious material is: slag powder 61.83%, steel slag powder 28.17%, and desulfurized gypsum 10%.

Images