An eps composite light-weight structure board and a production method thereof

Abstract

The application provides an EPS composite light structure plate and a production method thereof, and belongs to the technical field of the EPS composite light structure plate.The application establishes a dynamic fluid topology evolution compensation algorithm, monitors torque fluctuation of a stirring motor in real time, calculates apparent viscosity of a slurry flow field based on a finite volume method, maps density distribution of EPS light aggregate particles, and outputs extrusion compensation, so as to actively correct particle segregation; the inside of a fiber cement plate is subjected to low-temperature normal-pressure plasma jet processing, and an interface modifier is sprayed to form an interface transition layer, so that the bonding strength of the core material layer and the panel is improved; high-frequency micro-amplitude vibration forming technology is combined with a mold prestress reverse compensation technology to control forming pressure, so as to inhibit warping deformation; and the technical problems that cement-based cementitious materials and EPS light aggregate particles are caused to float and segregate due to a large difference in density, and the physical and mechanical properties of the cross section of the formed plate are not uniform, are solved.

Description

Technical Field

[0001] This invention belongs to the technical field of EPS composite lightweight structural panels, and more specifically, relates to an EPS composite lightweight structural panel and its production method. Background Technology

[0002] EPS composite lightweight structural board is a composite board with fiber cement board as the face panel and cement-based EPS lightweight aggregate mixture as the core material. It is widely used in building interior partitions, exterior wall insulation, and floor slabs. Traditional production processes rely on manual experience to prepare the slurry, and the molding is completed through fixed-speed stirring, static mold pouring, and natural curing. The production process lacks real-time monitoring and feedback control of the slurry flow field.

[0003] In the aforementioned traditional processes, the particle density of EPS lightweight aggregate is typically between 10 and 30. The density of cement-based binders is between 1500 and 1800. The density difference between the two is as high as tens to hundreds of times, causing the EPS lightweight aggregate particles to continuously migrate and accumulate upwards under buoyancy during slurry mixing and casting. This segregation effect is further aggravated during the molding vibration stage. Traditional processes use methods such as increasing the water-cement ratio or increasing the mixing time to suppress segregation. However, increasing the water-cement ratio will significantly reduce the strength of the core layer, and increasing the mixing time is difficult to determine the termination time due to the lack of quantitative characterization methods for flow field uniformity. Ultimately, it is still impossible to eliminate the problem of uneven particle density at the mechanistic level.

[0004] In existing technologies, due to the significant density difference between cement-based cementitious materials and EPS lightweight aggregate particles, the EPS lightweight aggregate particles continuously migrate upwards under buoyancy during mixing and casting, resulting in a stratified state where the upper part of the cross-section is enriched with particles and the lower part is sparse. This leads to uneven distribution of compressive strength across the entire cross-section of the molded board and a decrease in thermal and sound insulation performance. Furthermore, existing processes lack real-time quantitative characterization and active compensation methods for the uniformity of the slurry flow field, making it difficult to ensure consistent product quality under mass production conditions. In other words, existing technologies suffer from the technical problem of particle segregation due to the significant density difference between cement-based cementitious materials and EPS lightweight aggregate particles, resulting in uneven physical and mechanical properties of the molded board cross-section. Summary of the Invention

[0005] In view of this, the present invention provides an EPS composite lightweight structural board and its production method, which can solve the technical problems in the prior art where the large difference in density between cement-based cementitious material and EPS lightweight aggregate particles leads to particle floating and segregation, and the physical and mechanical properties of the formed board cross section are uneven.

[0006] This invention is achieved as follows: It provides an EPS composite lightweight structural panel and its production method, comprising fiber cement board, an interface transition layer, and a core material layer. The fiber cement board is symmetrically arranged on both sides of the EPS composite lightweight structural panel. The interface transition layer is located between the inner side of each fiber cement board and the core material layer, formed by spraying epoxy resin emulsion or polymer-modified cement slurry onto the inner surface of the fiber cement board after low-temperature, normal-pressure plasma jet treatment. This bonding process combines chemical bonding and physical anchoring to combine the core material layer with the fiber cement board. The core material layer is located between two fiber cement boards and is formed by uniformly mixing cement-based binder, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent, and multidimensional flexible fibers before casting. The multidimensional flexible fibers form a three-dimensional spatial network anchoring structure within the core material layer, which contains closed microbubbles introduced by the air-entraining agent. The edges of the EPS composite lightweight structural panel utilize a tongue-and-groove structure.

[0007] The fiber cement board is specifically formed by silicate cement, quartz sand and cellulose fibers through a flow casting or pressing method. It does not contain asbestos or halogen components, and the inner surface is treated with low-temperature and normal-pressure plasma jet to introduce polar functional groups.

[0008] Specifically, the EPS lightweight aggregate particles are expanded polystyrene foam particles with a particle size range of 1–5 mm and a density range of 10–30. The volumetric doping level ranges from 15% to 40%; the modified cellulose ether is a product of hydrophobic modification of hydroxypropyl methylcellulose ether; the multidimensional flexible fiber is at least two of polypropylene fiber, basalt fiber, and polyvinyl alcohol fiber, with a doping level ranging from 0.8% to 1.5%. .

[0009] This includes the following steps:

[0010] Cement-based cementitious material, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent and multidimensional flexible fiber are added to the mixer according to the formula. The dynamic fluid topology evolution compensation algorithm is started to collect the torque fluctuation data of the mixing motor in real time. The mixing speed is adjusted according to the output result of the dynamic fluid topology evolution compensation algorithm to obtain a uniform slurry.

[0011] The inner side of the fiber cement board is subjected to low-temperature atmospheric pressure plasma jet treatment. After the treatment is completed, an interface modifier is pre-sprayed on the inner side of the fiber cement board to form an interface transition layer.

[0012] The uniform slurry is injected into a mold made of fiber cement board that has undergone low-temperature and normal-pressure plasma jet treatment. The high-frequency micro-amplitude vibration molding process is used, and the molding pressure is controlled by the mold prestress reverse compensation technology. The molded board is obtained by one-time composite molding.

[0013] The multi-source heterogeneous feature correlation constraint evolution algorithm is activated to collect temperature sensor data, humidity sensor data and shrinkage sensor data in the curing environment in real time. Based on the output results of the multi-source heterogeneous feature correlation constraint evolution algorithm, the gradient curing process is executed. The relative humidity of the curing environment is precisely controlled in stages to complete the hardening of the molded board and obtain the hardened board.

[0014] The hardened boards are tested sequentially for surface flatness deviation, diagonal difference, compressive strength, sound insulation and fire resistance limit. Hardened boards that pass all tests are then stacked and stored.

[0015] Specifically, the dynamic fluid topology evolution compensation algorithm is based on the principle of the finite volume method. It establishes a coupled dynamic equation between the uniform slurry continuous medium and EPS lightweight aggregate particles, calculates the apparent viscosity of the flow field in real time, maps the density distribution probability map of EPS lightweight aggregate particles using coordinate transformation, and outputs the extrusion compensation amount at the outlet of the molding machine based on the particle density deviation.

[0016] Specifically, the torque fluctuation threshold range and the compression compensation adjustment coefficient in the dynamic fluid topology evolution compensation algorithm are determined by systematically changing the water-cement ratio, EPS lightweight aggregate particle volume content, and stirring speed through no less than 30 sets of stirring molding experiments, with the goal of minimizing the standard deviation of cross-sectional EPS lightweight aggregate particle density through multiple iterative fitting.

[0017] Specifically, the low-temperature atmospheric pressure plasma jet treatment refers to an energy density range of 5–15. It is used to remove organic pollutants from the inner surface of fiber cement board and introduce polar functional groups; the interface modifier is epoxy resin emulsion or polymer modified cement slurry, and the thickness of the interface transition layer ranges from 0.3 to 0.8 mm.

[0018] The high-frequency micro-amplitude vibration molding process specifically has a vibration frequency range of 50-100Hz and an amplitude range of 0.1-0.5mm. The vibration frequency range and amplitude range are determined by testing the content of closed micro-bubbles in the cross-section of the molded sheet and the uniformity of EPS lightweight aggregate particle distribution under different frequency and amplitude combinations.

[0019] Specifically, the mold prestressing reverse compensation technology involves applying an elastic pre-deformation amount, opposite to the direction of shrinkage deformation, to the bottom plate and side plates of the mold in the mold design stage, based on the drying shrinkage law of inorganic materials. The elastic pre-deformation amount is calculated according to the formula... calculate.

[0020] The gradient curing process specifically divides the hardening process into three stages: initial, intermediate, and final. The relative humidity is controlled within the range of 90% to 95% in the initial stage, 75% to 85% in the intermediate stage, and 55% to 65% in the final stage. The duration of each stage is adjusted according to the thickness of the formed board.

[0021] Specifically, the multi-source heterogeneous feature correlation constraint evolution algorithm combines the minimum spanning tree principle in graph theory to perform feature dimensionality reduction. It regards the sensor parameters corresponding to temperature sensor data, humidity sensor data, and shrinkage sensor data as nodes in the graph. The edge weights between nodes are defined by mutual information. Through a dynamic evolution mechanism, it continuously prunes redundancy and low-correlation connections, while retaining the core path.

[0022] The mutual information quantity is specifically calculated according to the formula. The probability distributions are calculated and estimated from historical production data.

[0023] The multi-source heterogeneous feature correlation constraint evolution algorithm also executes the Lagrange multiplier method on the core path to solve the parameter optimal domain under multi-objective constraints and outputs the relative humidity adjustment command of the curing environment to achieve a Pareto optimal balance between energy consumption and the pass rate of hardened boards.

[0024] The duration of each stage in the gradient curing process is specifically adjusted according to the thickness of the molded board within the range of 50-200mm. For every 25mm increase in thickness, the duration of each stage is extended by 2-4 hours.

[0025] The qualified standards for the compressive strength test are not less than 3.5 MPa and the softening coefficient is not less than 0.8; the qualified standards for the sound insulation test are a single value evaluation of not less than 40 dB; the qualified standards for the fire resistance limit test are not less than 1.5 h for the 50 mm and 60 mm series, not less than 3 h for the 75 mm, 90 mm and 100 mm series, and not less than 4 h for the 120 mm, 150 mm and 200 mm series; the flatness deviation of the board surface does not exceed 5 mm and the diagonal difference does not exceed 6 mm.

[0026] This invention introduces a dynamic fluid topology evolution compensation algorithm to collect torque fluctuation data of the stirring motor in real time. Based on the finite volume method, it establishes a coupled dynamic equation between the slurry and EPS lightweight aggregate particles, calculates the apparent viscosity of the current flow field and maps the particle density distribution probability, and automatically outputs the extrusion compensation amount at the outlet of the molding machine according to the particle density deviation, thereby realizing active real-time correction of particle segregation.

[0027] The above method establishes a quantitative characterization mechanism for the uniformity of the slurry flow field during the mixing stage, which completely changes the passive mode of traditional processes that rely on human experience to judge the end point of mixing. This ensures that the slurry is always in a controlled and uniform distribution state in three-dimensional space, and eliminates the buoyancy-driven migration effect caused by the large density difference of EPS lightweight aggregate particles from the mechanism level.

[0028] In summary, this invention solves the technical problem mentioned in the background art, which is that the large difference in density between cement-based cementitious materials and EPS lightweight aggregate particles leads to particle floating and segregation, and uneven physical and mechanical properties of the molded board cross-section. Attached Figure Description

[0029] Figure 1 This is a flowchart of the method of the present invention.

[0030] Figure 2 This is a probability evolution diagram of the cross-sectional density distribution of EPS lightweight aggregate particles at different mixing stages.

[0031] Figure 3 The graph shows the time sequence control curves of relative humidity in the three-stage curing room for the gradient curing process.

[0032] Figure 4 This is a histogram showing the volume fraction distribution of EPS lightweight aggregate particles along the height of the hardened board cross-section. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below.

[0034] like Figure 1 The diagram shown is a flowchart of an EPS composite lightweight structural panel and its production method provided by the present invention. The method includes the following steps:

[0035] S01. Add cement-based cementitious material, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent and multidimensional flexible fiber to the mixer according to the proportion, start the dynamic fluid topology evolution compensation algorithm, collect the torque fluctuation data of the mixing motor in real time, and adjust the mixing speed according to the output result of the dynamic fluid topology evolution compensation algorithm to obtain a uniform slurry.

[0036] S02. The inner side of the fiber cement board is subjected to low-temperature atmospheric pressure plasma jet treatment. After the treatment is completed, an interface modifier is pre-sprayed on the inner side of the fiber cement board to form an interface transition layer. The thickness of the interface transition layer is 0.3 to 0.8 mm.

[0037] S03. The uniform slurry obtained in step S01 is injected into the mold formed by the fiber cement board processed in step S02. A high-frequency micro-amplitude vibration molding process with a frequency range of 50-100Hz and an amplitude range of 0.1-0.5mm is adopted. The molding pressure is controlled by the mold prestress reverse compensation technology. The molded board is obtained by one composite molding.

[0038] S04. Start the multi-source heterogeneous feature correlation constraint evolution algorithm, collect temperature sensor data, humidity sensor data and shrinkage sensor data in the curing environment in real time, execute the gradient curing process according to the output result of the multi-source heterogeneous feature correlation constraint evolution algorithm, accurately control the relative humidity of the curing environment in stages, complete the hardening of the molded board, and obtain the hardened board.

[0039] S05. The hardened boards are tested in sequence for flatness deviation, diagonal difference, compressive strength, sound insulation and fire resistance limit. Hardened boards that pass all tests are put into the stacking and storage process.

[0040] The EPS lightweight aggregate particles are expanded polystyrene foam particles with a particle size range of 1–5 mm and a density range of 10–30. The volumetric doping range is 15% to 40%; the energy density range of the low-temperature atmospheric pressure plasma jet treatment is 5 to 15. .

[0041] Among them, the modified cellulose ether is the product of hydrophobic modification of hydroxypropyl methylcellulose ether. Its function is to adjust the thixotropy of the uniform slurry. Thixotropy refers to the rheological characteristics of the uniform slurry, which decreases in viscosity when subjected to shear and increases in viscosity when left to stand. This allows it to maintain fluidity during stirring and solidify rapidly after molding, thus inhibiting the floating and segregation of EPS lightweight aggregate particles.

[0042] Among them, the air-entraining agent is a rosin soap or alkyl sulfonate surfactant, which is used to introduce closed microbubbles into the homogeneous slurry. The diameter of the closed microbubbles ranges from 50 to 300 μm, and it works synergistically with the modified cellulose ether to adjust the apparent viscosity of the homogeneous slurry.

[0043] The multidimensional flexible fiber is at least two of polypropylene fiber, basalt fiber, and polyvinyl alcohol fiber, with a fiber length ranging from 6 to 19 mm and a dosage ranging from 0.8 to 1.5%. Multidimensional flexible fibers form a three-dimensional spatial network anchoring structure within the hardened core material layer, alleviating the shear stress caused by the difference in thermal expansion coefficients at the interface between the core material layer and the fiber cement board.

[0044] The dynamic fluid topology evolution compensation algorithm, based on the finite volume method, establishes a coupled dynamic equation between the uniform slurry continuous medium and EPS lightweight aggregate particles. It monitors the torque fluctuation data of the stirring motor in real time and calculates the current apparent viscosity of the flow field. Using coordinate transformation, it maps the probability map of the EPS lightweight aggregate particle density distribution inside the board. Based on the EPS lightweight aggregate particle density deviation, it automatically calculates and outputs the extrusion compensation amount at the molding machine outlet. Through reverse fluid dynamics correction, it corrects the product density unevenness caused by EPS lightweight aggregate particle segregation in real time. This dynamic fluid topology evolution compensation algorithm ensures that the uniform slurry flow field is in a controlled and uniform distribution state in three-dimensional space. It fundamentally solves the problem of particle floating and segregation caused by the significant density difference between cement-based cementitious materials and EPS lightweight aggregate particles, ensuring the consistency of the physical and mechanical properties of the entire cross-section of the molded board. It avoids the problems of uneven strength distribution and decreased thermal and sound insulation performance caused by local enrichment or depletion of EPS lightweight aggregate particles, significantly improving the quality stability of the hardened board.

[0045] Among them, the torque fluctuation threshold range and the compression compensation adjustment coefficient used in the dynamic fluid topology evolution compensation algorithm were obtained through no less than 30 sets of stirring molding experiments. In the experiments, the water-cement ratio (range 0.35 to 0.55), the volume content of EPS lightweight aggregate particles (range 15% to 40%) and the stirring speed (range 20 to 80 r / min) were systematically changed. After each set of specimens was cut according to the standard, the cross-sectional EPS lightweight aggregate particle density distribution was measured. With the minimum standard deviation of cross-sectional EPS lightweight aggregate particle density as the objective, the optimal domain of each control parameter was determined through multiple iterative fitting.

[0046] The energy density range of low-temperature atmospheric pressure plasma jet treatment is 5–15. It is used to remove organic pollutants from the inner surface of fiber cement board and introduce polar functional groups. The energy density range is determined by interfacial peel strength test experiments after treatment with different energy densities, with the energy density range corresponding to the peak value of interfacial peel strength as the standard.

[0047] Among them, the interface modifier is epoxy resin emulsion or polymer modified cement slurry, the thickness of the interface transition layer is 0.3 to 0.8 mm, the interface transition layer penetrates into the core material layer, and the core material layer and fiber cement board are tightly bonded through the dual action of chemical bonding and physical anchoring, and the interface peel strength is not less than 0.4 MPa.

[0048] Among them, the vibration frequency range in the high-frequency micro-amplitude vibration molding process is 50-100Hz and the amplitude range is 0.1-0.5mm. The vibration frequency range and amplitude range are determined by testing the content of closed microbubbles in the cross section of the molded sheet and the uniformity of EPS lightweight aggregate particle distribution under different frequency amplitude combinations. The parameter range with the lowest content of closed microbubbles and the most uniform distribution of EPS lightweight aggregate particles is selected.

[0049] Among them, the mold prestress reverse compensation technology refers to applying an elastic pre-deformation amount opposite to the shrinkage deformation direction to the bottom plate and side plate of the mold in the mold design stage, based on the drying shrinkage law of inorganic materials. This allows the free shrinkage of the molded sheet after demolding to cancel out the elastic pre-deformation amount, thereby suppressing the warping deformation caused by uneven water loss shrinkage of large-size molded sheets during the hardening process.

[0050] The calculation formula for the elastic pre-deformation in the mold prestressing reverse compensation technology is expressed as follows:

[0051] ;

[0052] In the formula, The elastic pre-deformation amount of the mold (mm). The nominal length (mm) of the formed sheet material. The shrinkage coefficient (mm / mm) is the core material shrinkage coefficient. This is the relative water loss rate (dimensionless). The asymmetric structure correction coefficient (dimensionless) is determined by regression analysis after weighing the water loss and measuring the bending deformation of the molded sheet in the gradient curing experiment.

[0053] The gradient curing process refers to dividing the hardening process into three stages: initial, middle, and final. The relative humidity is controlled within the range of 90% to 95% in the initial stage, 75% to 85% in the middle stage, and 55% to 65% in the final stage. The duration of each stage is adjusted according to the thickness of the molded board (50 to 200 mm). For every 25 mm increase in thickness, the duration of each stage is extended by 2 to 4 hours. The duration of each stage is determined by testing the shrinkage rate of molded boards of different thicknesses under various relative humidities.

[0054] The multi-source heterogeneous feature correlation constraint evolution algorithm addresses the problem of sensor parameter dimensionality explosion during production. It combines the minimum spanning tree principle from graph theory for feature dimensionality reduction, treating the sensor parameters corresponding to temperature, humidity, and shrinkage sensor data as nodes in a graph. Edge weights between nodes are defined by mutual information. A dynamic evolution mechanism continuously prunes redundancy and low-correlation connections, preserving core paths sensitive to the quality of the hardened board. Subsequently, the Lagrange multiplier method is executed on the core paths to solve for the parameter optimal domain under multi-objective constraints, and a relative humidity adjustment command for the curing environment is output. This multi-source heterogeneous feature correlation constraint evolution algorithm enables the production control system to automatically filter out core process parameters that significantly affect the sound insulation, compressive strength, and surface flatness deviation of the hardened board from temperature, humidity, and shrinkage sensor data, eliminating redundant interference signals. While ensuring production speed, it achieves a Pareto optimal balance between energy consumption and the pass rate of the hardened board, thereby maintaining a high degree of consistency in the quality of the hardened board under large-scale continuous production conditions.

[0055] Mutual information is an indicator that measures the degree of statistical correlation between two random variables. The formula for calculating mutual information is as follows:

[0056] ;

[0057] In the formula, For variables With variables Mutual information (bits) between them. For variables Information entropy (bit) Let be the joint probability (dimensionless). and These are the marginal probabilities (dimensionless), and each probability distribution is estimated statistically from historical production data.

[0058] Pareto optimal equilibrium refers to a multi-objective optimization problem in which no single solution can further improve the state of a particular objective without harming other objectives; that is, multiple objectives simultaneously reach the optimal overall equilibrium point.

[0059] Among them, the Lagrange multiplier method refers to a mathematical method for solving the extrema of a multi-objective function under constraints. By introducing Lagrange multipliers, the constraints are incorporated into the objective function, which is then transformed into an unconstrained optimization problem for solution.

[0060] In step S05, the pass standards for compressive strength testing are not less than 3.5 MPa, softening coefficient not less than 0.8, sound insulation testing is not less than 40 dB for a single value evaluation, and fire resistance limit testing is not less than 1.5 h for 50 mm and 60 mm series, not less than 3 h for 75 mm, 90 mm and 100 mm series, and not less than 4 h for 120 mm, 150 mm and 200 mm series. The flatness deviation of the board surface shall not exceed 5 mm, and the diagonal difference shall not exceed 6 mm. All testing methods shall be carried out in accordance with the current national standards for testing building materials.

[0061] The detailed structure of an EPS composite lightweight structural panel is as follows:

[0062] EPS composite lightweight structural board consists of three parts: fiber cement board, interface transition layer, and core material layer. It is manufactured through a one-time composite molding process. The shape is a rectangular board with a nominal length of 2440mm or 3000mm, a nominal width of 610mm, and nominal thicknesses of 50mm, 60mm, 75mm, 90mm, 100mm, 120mm, 150mm, and 200mm. The allowable tolerances are ±5mm in length, ±5mm in width, and ±5mm in thickness. The flatness deviation of the board surface shall not exceed 5mm, and the diagonal difference shall not exceed 6mm.

[0063] Fiber cement board is located on both sides of EPS composite lightweight structural board, with one fiber cement board on each side. The two fiber cement boards are arranged symmetrically. The thickness of a single fiber cement board ranges from 4 to 8 mm. The fiber cement board is formed by silicate cement, quartz sand and cellulose fiber through a slurry method or a pressing method. The fiber cement board does not contain asbestos or halogen components. The inner surface of the fiber cement board is treated with low temperature and normal pressure plasma jet to introduce polar functional groups. The single-point static hanging load is not less than 60 kg. The fiber cement board is suitable for directly bonding tiles, marble and decorative panels, and is also suitable for directly nailing sanitary ware, cabinets and pipe supports.

[0064] The interface transition layer is located between the inner side of each fiber cement board and the core material layer. The thickness of the interface transition layer ranges from 0.3 to 0.8 mm. The interface transition layer is formed by uniformly spraying epoxy resin emulsion or polymer-modified cement slurry onto the surface of the inner side of the fiber cement board after low-temperature and normal-pressure plasma jet treatment. The interface transition layer penetrates into the core material layer side and tightly bonds the core material layer and the fiber cement board through the dual effects of chemical bonding and physical anchoring. The interface peel strength is not less than 0.4 MPa.

[0065] The core layer is located between two fiber cement boards. It is formed by casting a uniform mixture of cement-based binder, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent, and multidimensional flexible fibers. The EPS lightweight aggregate particle size ranges from 1 to 5 mm, and the EPS lightweight aggregate particle density ranges from 10 to 30. The volumetric content of EPS lightweight aggregate ranges from 15% to 40%, and the EPS lightweight aggregate particles are uniformly dispersed in the cementitious matrix without stratification or particle enrichment. Multidimensional flexible fibers form a three-dimensional spatial network anchoring structure within the core layer, with a content ranging from 0.8% to 1.5%. The core material layer contains closed microbubbles introduced by an air-entraining agent, with a diameter ranging from 50 to 300 μm; the compressive strength of the core material layer is not less than 3.5 MPa, and the softening coefficient is not less than 0.8; the surface density of the EPS composite lightweight structural board ranges from 40 to 115. The fire resistance of EPS composite lightweight structural panels increases with the increase of nominal thickness; the fire resistance limit of EPS composite lightweight structural panels is 1.5 to 4 hours corresponding to the nominal thickness series, and they do not burn or release toxic or harmful gases at a high temperature of 1000℃; the sound insulation of EPS composite lightweight structural panels ranges from 40 to 50 dB.

[0066] The EPS composite lightweight structural panel adopts a tongue and groove structure on all four sides. The tongue and groove are processed on one side of the long side of the EPS composite lightweight structural panel and the tongue and groove protrusion is processed on the other side. When adjacent EPS composite lightweight structural panels are installed, the tongue and groove protrusion is inserted into the tongue and groove to form a seamless splice. The splice gap is filled with polymer modified mortar to ensure the continuity of sound insulation performance, fire resistance limit and waterproof performance at the splice.

[0067] The specific implementation of step S01 is as follows: Cement-based cementitious material, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent, and multidimensional flexible fiber are sequentially added to a mixer according to a predetermined ratio. With a water-cement ratio of 0.35–0.55 and an EPS lightweight aggregate particle volume fraction of 15%–40% as parameters, a dynamic fluid topology evolution compensation algorithm is initiated. This algorithm divides the mixing chamber into discrete control volumes based on the finite volume method, establishes a coupled dynamic equation between the uniform slurry continuous medium and the EPS lightweight aggregate particles, and uses the torque fluctuation data of the mixing motor as input to calculate the local apparent viscosity of each control volume in real time. Then, through coordinate transformation, the local apparent viscosity is mapped to a density distribution probability map of the EPS lightweight aggregate particles in three-dimensional space. When the standard deviation of particle density at a certain cross-section exceeds a set threshold, the algorithm automatically calculates the extrusion compensation amount at the molding machine outlet and issues a speed adjustment command. Through reverse fluid dynamics correction, the buoyancy driving force is offset, ensuring that the EPS lightweight aggregate particles remain uniformly dispersed, ultimately producing a uniform slurry. The torque fluctuation threshold and the compression compensation adjustment coefficient have been calibrated through no less than 30 sets of experiments. The reference value is the parameter range corresponding to when the standard deviation of cross-sectional particle density does not exceed 8%.

[0068] The specific implementation of step S02 is as follows: the inner side of the fiber cement board is placed under a low-temperature, normal-pressure plasma jet device with an energy density of 5-15 Surface activation treatment is performed. A low-temperature, ambient-pressure plasma jet generates a highly active particle beam under ambient pressure, bombarding the inner surface of the fiber cement board. This removes organic contaminants, breaks surface chemical bonds, and introduces polar functional groups such as hydroxyl and carboxyl groups, significantly improving surface wettability. Immediately after plasma jet treatment, an epoxy resin emulsion or polymer-modified cement slurry is uniformly sprayed onto the treated inner surface using a pneumatic atomizing spray gun, forming an interfacial transition layer with a thickness of 0.3–0.8 mm. One end of the polymer molecular chain in the interfacial transition layer chemically bonds to the fiber cement board surface through polar functional groups, while the other end penetrates into the core material layer to form a physical anchor, thus achieving a macroscopic bonding effect with an interfacial peel strength of not less than 0.4 MPa.

[0069] The specific implementation of step S03 is as follows: Two fiber cement boards processed in step S02 are symmetrically placed in a molding mold to form a closed mold cavity, and the uniform slurry obtained in step S01 is injected into the mold cavity. The molding stage employs a high-frequency micro-amplitude vibration molding process, with the vibration frequency set in the range of 50–100 Hz and the amplitude set in the range of 0.1–0.5 mm. The vibration energy is sufficient to allow the slurry to flow and fill fully without generating inertial forces sufficient to drive the macroscopic migration of EPS lightweight aggregate particles, thereby promoting the escape of air bubbles while maintaining uniform particle distribution. The mold prestressing reverse compensation technology is based on the formula during the mold design stage. An elastic pre-deformation amount is applied to the mold base plate and side plates in advance, wherein the shrinkage coefficient is... Relative water loss rate and asymmetric structure correction coefficient All were determined through gradient curing experiments and regression analysis. The elastic pre-deformation amount and the free shrinkage amount after demolding of the molded sheet cancel each other out, suppressing the warping deformation caused by uneven shrinkage due to water loss during the hardening process of large-sized sheets, completing one composite molding and obtaining the molded sheet.

[0070] The specific implementation of step S04 is as follows: the molded sheet material is sent into the intelligent curing chamber, and the multi-source heterogeneous feature correlation constraint evolution algorithm is started. This algorithm treats the sensor parameters corresponding to the temperature sensor data, humidity sensor data, and shrinkage sensor data as nodes in a graph theory, with the edge weights between nodes determined by normalized mutual information. The process involves defining a minimum spanning tree principle for feature dimensionality reduction, dynamically pruning redundant connections with mutual information below a set threshold (reference value 0.15), and retaining core path nodes that significantly affect the sound insulation, compressive strength, and surface flatness deviation of the hardened board. Subsequently, the Lagrange multiplier method is executed on the core path, using the compressive strength and pass rate of the hardened board as objective functions and energy consumption as a constraint, to solve for the optimal relative humidity domain under multi-objective constraints, outputting relative humidity adjustment commands for each stage of curing. The gradient curing process divides the hardening process into three stages: initial, intermediate, and final. Initially, the relative humidity is maintained at 90%–95% to ensure the cement hydration rate and prevent early water loss and cracking; in the intermediate stage, the relative humidity decreases to 75%–85% to promote the strength development of the core layer; and in the final stage, the relative humidity further decreases to 55%–65% to accelerate board drying and inhibit later shrinkage. The duration of each stage is adjusted according to the thickness of the formed board; for every 25mm increase in thickness, the duration of each stage is extended by 2–4 hours to complete the hardening of the formed board, resulting in a hardened board.

[0071] The specific implementation method of step S05 is as follows: All quality inspections are performed on the hardened boards sequentially. The flatness deviation of the board surface is measured using a straightedge and feeler gauge, with a passing standard of no more than 5mm; the diagonal difference is measured using a steel tape measure, with a passing standard of no more than 6mm; the compressive strength of the core material layer is determined using a pressure testing machine, with a passing standard of no less than 3.5MPa and a softening coefficient of no less than 0.8; the sound insulation is determined using the reverberation chamber method or microphone array method, with a passing standard of a single-value evaluation of no less than 40dB; and furnace combustion tests are conducted according to the fire resistance limit standards for different thickness series: no less than 1.5h for 50mm and 60mm thicknesses, no less than 3h for 75mm, 90mm, and 100mm thicknesses, and no less than 4h for 120mm, 150mm, and 200mm thicknesses. Only hardened boards that pass all inspections can enter the stacking and storage process. Boards that fail any inspection are deemed unqualified and are isolated according to regulations.

[0072] It should be noted that the key technologies of this invention include: a dynamic fluid topology evolution compensation algorithm based on the finite volume method to achieve real-time analysis of the torque fluctuation of the stirring motor, transforming the unobservable particle distribution state inside the slurry into a quantifiable density deviation, and actively applying compression compensation through reverse fluid dynamics correction, thus eliminating the buoyancy-driven migration of EPS lightweight aggregate particles at the mechanistic level; a synergistic interface engineering of low-temperature atmospheric pressure plasma jet treatment and interface modifier spraying, which improves the interface peeling strength to a level that is difficult to achieve by traditional mechanical roughening processes through the dual mechanisms of polar functional group introduction and polymer penetration anchoring; and a multi-source heterogeneous feature correlation constraint evolution algorithm that solves the problem by combining minimum spanning tree pruning and the Lagrange multiplier method, automatically selecting core process parameters in a production environment with an explosion of sensor parameter dimensions, and achieving precise closed-loop control of the gradient curing process. The synergistic effect of the above three key technologies enables the three links of mixing uniformity control, interface bonding strength assurance and curing process precise regulation to form a complete closed loop in a logical manner. The precise control of any link is based on the stable output of other links, thus ensuring the overall quality consistency of EPS composite lightweight structural panels from raw materials to finished products.

[0073] It should be noted that under conditions of large-scale continuous production, when molded boards of different thicknesses and with different entry times are present in the curing chamber, a single fixed humidity curing regime cannot simultaneously meet the hydration requirements of all boards at each stage. Traditional methods circumvent this contradiction by extending the uniform curing cycle, but this leads to over-drying of thin boards and insufficient hydration of thick boards, significantly increasing batch-to-batch quality dispersion. The reason for this technical problem is that the moisture diffusion path lengths of boards of different thicknesses differ by orders of magnitude, resulting in a time span required for each board to reach the same degree of hydration that is far greater than the adjustment margin that a single curing regime can accommodate. Furthermore, the curing chamber contains a large number of temperature and humidity sensors with extremely high data dimensionality, making it impossible for manual identification in real time of which sensor parameters have the most significant impact on the quality of boards of specific thicknesses, thus hindering targeted adjustments to the local curing environment. A common solution to this technical problem is to set up independent curing chambers for different thickness series, configuring a separate curing control system for each thickness range. However, this method significantly increases equipment investment and space occupation, and when product models are frequently switched, the boundary conditions of each curing zone are still difficult to match precisely, and the problem of board quality fluctuations at the boundaries of different zones persists. This invention effectively solves this technical problem. The multi-source heterogeneous feature correlation constraint evolution algorithm uses the minimum spanning tree principle of graph theory to dynamically reduce the dimensionality of all sensor parameters. It identifies and retains core sensor nodes sensitive to the quality of the current batch of boards using mutual information as edge weights, while eliminating redundant interference signals. Subsequently, the Lagrange multiplier method is executed on the core path to solve the optimal relative humidity domain in real time with the board compressive strength and pass rate as objective functions, outputting precise humidity adjustment instructions for the actual board combination in the current curing chamber. This mechanism enables a single curing chamber to dynamically adjust curing parameters according to the real-time status of boards of different thicknesses without adding zoning equipment. It achieves a Pareto optimal balance between energy consumption and the pass rate of hardened boards while ensuring production speed, thereby maintaining a high degree of consistency in the quality of hardened boards under large-scale continuous production conditions.

[0074] Specifically, the principle of this invention is:

[0075] The fundamental reason why this invention can solve the above-mentioned technical problems is that the dynamic fluid topology evolution compensation algorithm indirectly characterizes the particle distribution state inside the slurry flow field, which cannot be directly observed, through the measurable external signal of the torque fluctuation of the stirring motor, thereby establishing a causal mapping chain from macroscopic measurability to microscopic particle distribution.

[0076] Specifically, the spatial distribution of EPS lightweight aggregate particles in the slurry determines the local apparent viscosity distribution of the mixing system. This spatial non-uniformity of the local apparent viscosity directly reflects the periodic fluctuations in the resistance experienced by the stirring blades at different locations, which in turn manifests as the temporal fluctuations in the output torque of the stirring motor. Based on the finite volume method, the coupled dynamic equations of the continuous medium and particles can be used to solve the aforementioned torque fluctuation signals into apparent viscosity values ​​at various spatial locations within the flow field. Then, through coordinate transformation, a three-dimensional density distribution probability map of the EPS lightweight aggregate particles can be mapped, allowing for the quantitative calculation of the particle density deviation at each cross-sectional location.

[0077] Based on this, the algorithm automatically calculates the extrusion compensation amount at the molding machine outlet according to the particle density deviation. Through reverse fluid dynamics correction, a pressure gradient opposite to the segregation direction is applied at the outlet to actively counteract the buoyancy driving force, allowing the particles to return to a uniform distribution state before the slurry solidifies. This active closed-loop compensation mechanism works synergistically with the thixotropic thickening effect provided by the modified cellulose ether. The former is responsible for correcting the particle distribution deviation in real time during the flow state, while the latter is responsible for rapidly increasing the static viscosity after molding to lock the particle position. Together, they suppress particle segregation from both dynamic and static dimensions.

[0078] Furthermore, the high-frequency micro-amplitude vibration molding process promotes slurry flow and filling without disrupting the uniformity of the already formed particle distribution through vibration frequencies of 50–100 Hz. The closed microbubbles introduced by the air-entraining agent further reduce the driving force for particle migration during the vibration stage by occupying the interparticle gaps and increasing the equivalent viscosity of the system. The gradient curing process controls the water loss rate of the molded board at each stage by gradually reducing the relative humidity of the environment, ensuring that the core layer maintains low shrinkage stress during the strength building period and avoiding localized cracking caused by rapid water loss that could disrupt the uniformity of particle distribution. Logically, these steps form a synergistic control system covering the entire process from mixing and molding to curing, ensuring that the uniform particle distribution is maintained throughout the hardening process and ultimately guaranteeing the consistency of the physical and mechanical properties of the entire cross-section of the molded board.

[0079] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0080] The specific implementation method of step S01 is as follows.

[0081] Cement-based binder, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent, and multidimensional flexible fibers are added to the mixer according to the specified proportions, and a dynamic fluid topology evolution compensation algorithm is initiated. This algorithm, based on the finite volume method, establishes a coupled dynamic equation between the homogeneous slurry continuous medium and the EPS lightweight aggregate particles. It simultaneously solves for the buoyancy, drag, and diffusion forces acting on the discrete phase particles. The coupled dynamic equation is expressed as follows:

[0082] ;

[0083] In the formula, The particle density of EPS lightweight aggregate ( (The range is 10 to 30) For the first The velocity of each particle ( ) For time ( ) For the first The buoyant force on each particle ( ) For the first The drag force on each particle ( ) For the first The diffusion force on each particle ( ) For the first The dynamic error term of each particle ( ).

[0084] buoyancy The formula is expressed as follows:

[0085] ;

[0086] In the formula, For the density of a uniform slurry continuous medium ( ) The acceleration due to gravity ( ), take 9.81 Volume of a single EPS lightweight aggregate particle ( ).

[0087] Traction The formula is expressed as follows:

[0088] ;

[0089] In the formula, The drag coefficient (dimensionless) is determined by the particle Reynolds number, with an empirical value of 0.4–0.8. The cross-sectional area of ​​a single EPS lightweight aggregate particle facing the flow ( ) For the continuous medium velocity of the slurry ( ).

[0090] Diffusion force The formula is expressed as follows:

[0091] ;

[0092] In the formula, Let be the Boltzmann constant, taken as 1.38 × J / K; The absolute temperature of the slurry (in K). The diameter of EPS lightweight aggregate particles (in meters) ranges from [value missing]. ; For the first The position coordinate components of each particle (in meters); Spatial coordinates EPS lightweight aggregate particle density (unit) ) Target mean density (units) ).

[0093] The system collects real-time torque fluctuation data of the stirring motor and calculates the current apparent viscosity of the flow field. The calculation formula is as follows:

[0094] ;

[0095] In the formula, for At any given time, the apparent viscosity of the flow field ( ) Initial reference viscosity ( ) The torque-viscosity conversion factor (dimensionless) is obtained from calibration experiments. for The torque of the stirring motor at all times ( ) Reference torque ( ) This is the viscosity correction error term (dimensionless).

[0096] The probability map of EPS lightweight aggregate particle density distribution inside the board is mapped using coordinate transformation. The coordinate transformation formula is expressed as follows:

[0097] ;

[0098] In the formula, Total number of particles For the first The radial basis function (dimensionless) of a particle is expressed by the following formula:

[0099] ;

[0100] In the formula, For the first The spatial coordinates of each particle ( ) For the first The diffusion radius of each particle ( The empirical value is 0.5 to 1.0 times the particle size of EPS lightweight aggregate. For the first Particle density ( ) This is the density distribution error term (dimensionless).

[0101] The extrusion compensation amount at the molding machine outlet is automatically calculated based on the density deviation of EPS lightweight aggregate particles. The formula is expressed as follows:

[0102] ;

[0103] In the formula, For the amount of compression compensation ( ) For reference molding pressure ( ) , The proportional and differential control coefficients (dimensionless) were obtained by fitting no fewer than 30 sets of mixing and molding experiments. The experiments systematically varied the water-cement ratio (range 0.35–0.55), the volumetric content of EPS lightweight aggregate particles (range 15%–40%), and the stirring speed (range 20–80). The objective was to minimize the standard deviation of the cross-sectional EPS lightweight aggregate particle density, which was determined iteratively. To compensate for the error term (dimensionless), the algorithm adjusts the stirring speed in real time based on the output results to produce a uniform slurry.

[0104] The specific implementation method of step S02 is as follows.

[0105] The inner side of the fiber cement board is subjected to low-temperature, ambient-pressure plasma jet treatment with an energy density range of 5–15. This process is used to remove surface organic contaminants and introduce polar functional groups. The energy density range is determined by interfacial peel strength tests after treatment with different energy densities, with the range corresponding to the peak peel strength as the standard. After treatment, an interfacial modifier (epoxy resin emulsion or polymer-modified cement slurry) is pre-sprayed onto the inside of the fiber cement board, forming a layer with a thickness of 0.3–0.8 mm. The interface transition layer.

[0106] The specific implementation method of step S03 is as follows.

[0107] The uniform slurry is injected into the mold at a frequency of 50–100. Amplitude 0.1–0.5 The high-frequency micro-amplitude vibration molding process determines the frequency and amplitude parameters by testing the content of closed microbubbles in the cross-section of the molded sheet and the uniformity of EPS lightweight aggregate particle distribution under different combinations. The parameter range with the lowest content of closed microbubbles and the most uniform particle distribution is selected. Combining mold prestress reverse compensation technology to control the molding pressure, the formula for calculating the elastic pre-deformation is as follows:

[0108] ;

[0109] In the formula, The elastic pre-deformation amount of the mold ( ) The nominal length of the formed sheet ( ) The shrinkage coefficient of the core layer material ( ) Water loss ( ) Initial water content ( ) Relative water loss rate (dimensionless) The asymmetric structure correction coefficient is dimensionless. Each parameter was determined through regression analysis after measuring water loss and bending deformation of the molded sheet at each stage in the gradient curing experiment. The molded sheet was obtained through a single composite molding process.

[0110] The specific implementation method of step S04 is as follows.

[0111] A multi-source heterogeneous feature correlation-constrained evolution algorithm is initiated to collect real-time data from temperature, humidity, and shrinkage sensors in the maintenance environment. This algorithm combines the minimum spanning tree principle for feature dimensionality reduction, treating each sensor parameter as a node in a graph. The edge weights between nodes are defined by mutual information, and the formula for calculating mutual information is as follows:

[0112] ;

[0113] In the formula, For variables With variables Mutual information between them ) For variables Information entropy ( The formula is expressed as follows:

[0114] ;

[0115] In the formula, For variables Value ,variable Value Joint probability at time (dimensionless) For variables Value marginal probability (dimensionless) For variables Value The marginal probabilities (dimensionless) are estimated from historical production data.

[0116] The algorithm continuously prunes redundancies and low-correlation connections through a dynamic evolution mechanism, retaining core paths sensitive to the quality of the hardened slab. Then, the Lagrange multiplier method is used on the core paths to solve for the parameter optimum domain under multi-objective constraints. The Lagrange multiplier method incorporates the constraints into the objective function, transforming it into an unconstrained optimization problem for solution. The multi-objective constrained optimization formula is expressed as follows:

[0117] ;

[0118] In the formula, Vector for regulating relative humidity in the maintenance environment (dimensionless) The overall objective function (dimensionless) is composed of the weighted sum of the normalized values ​​of energy consumption and the normalized values ​​of the pass rate. , For Lagrange multipliers (dimensionless) The mass constraint function (dimensionless) is formed by the ratio of the measured sound insulation value to the acceptable threshold. The mass constraint function (dimensionless) is the ratio of the measured compressive strength to the acceptable threshold. The algorithm outputs a relative humidity adjustment command for the curing environment, executing a gradient curing process. The initial relative humidity control range is 90%–95%, the middle stage is 75%–85%, and the final stage is 55%–65%. The duration of each stage is adjusted according to the thickness of the formed sheet, with each increase in thickness by 25%. The duration of each stage is extended by 2-4 hours. The hardening process is completed to obtain hardened boards.

[0119] The specific implementation method of step S05 is as follows.

[0120] The hardened boards were tested sequentially for surface flatness deviation, diagonal difference, compressive strength, sound insulation, and fire resistance limit. The acceptable compressive strength standard was not less than 3.5. The softening coefficient is not less than 0.8, and the sound insulation qualification standard is a single-value evaluation quantity of not less than 40. The fire resistance limit qualification standard corresponds to 1.5 to 4 mm according to the thickness series. The flatness deviation of the board surface shall not exceed 5. The difference between the diagonals does not exceed 6 All testing methods are carried out in accordance with the current national standards for testing building materials. Hardened boards that pass all tests are then put into the stacking and storage process.

[0121] To better understand and implement this invention, the following is a specific application scenario of the invention, Example 2: In order to verify the effect of the invention, the technicians set up a test environment and produced a batch of EPS composite lightweight structural boards with a nominal thickness of 100mm, a nominal length of 2440mm, and a nominal width of 610mm. The technical parameters of each step of the invention and the performance of the final product were systematically recorded and analyzed.

[0122] In this production, 42.5 grade ordinary Portland cement was used as the cement-based binder, and the EPS lightweight aggregate particles were 3mm in size and 20 in density. The volumetric admixture was set at 30%, the water-cement ratio at 0.45, and the multidimensional flexible fiber was a combination of polypropylene fiber and basalt fiber, with an admixture content of 1.2%. The fiber length is 12mm. During the stirring stage, the dynamic fluid topology evolution compensation algorithm is activated. The system collects the torque fluctuation data of the stirring motor every 0.5s, and the initial stirring speed is set to 40r / min.

[0123] During the mixing process, the dynamic fluid topology evolution compensation algorithm continuously calculates the apparent viscosity distribution of the flow field and maps the calculation results to a three-dimensional density distribution probability map of EPS lightweight aggregate particles, such as... Figure 2 As shown, Figure 2 This reflects the probabilistic evolution of the density distribution of EPS lightweight aggregate particles along the cross-sectional height at different mixing stages. In the initial stage, the standard deviation of cross-sectional particle density was 21.3%, exceeding the set threshold of 8%. The algorithm automatically increased the mixing speed from 40 r / min to 62 r / min and output a compression compensation of 3.2 mm. After reverse fluid dynamics correction, the standard deviation of cross-sectional particle density decreased to 5.6% by the 8th minute of mixing, meeting the qualified standard, and the preparation of uniform slurry was completed.

[0124] During the interface treatment stage, the inner side of the fiber cement board is subjected to an energy density of 10 After low-temperature, atmospheric-pressure plasma jet treatment, the surface contact angle decreased from approximately 72° to approximately 18°, indicating that polar functional groups had been fully introduced. Subsequently, polymer-modified cement slurry was sprayed, with the interfacial transition layer thickness controlled at 0.5 mm. The measured interfacial peel strength was 0.63 MPa, meeting the qualified standard of not less than 0.4 MPa.

[0125] During the molding stage, the vibration frequency is set to 75Hz and the amplitude is set to 0.3mm. In the mold prestressing reverse compensation technology, the elastic pre-deformation is based on the shrinkage coefficient. =0.0004, relative water loss rate =0.18, asymmetric structure correction coefficient =1.12 Calculation, elastic pre-deformation =0.197mm, and this pre-deformation amount is applied to the mold base plate in advance. After pouring and vibration molding, the flatness deviation of the plate surface is checked after demolding. The deviation is 1.8mm and the diagonal difference is 2.3mm, both of which meet the qualified standards.

[0126] During the curing stage, molded sheets of three thicknesses—50mm, 100mm, and 150mm—were present in the curing chamber. A multi-source heterogeneous feature correlation constraint evolution algorithm was used to perform feature dimensionality reduction on data from 24 temperature and humidity sensors and 8 shrinkage sensors. The calculation results of the mutual information of each sensor parameter are shown in Table 1.

[0127] Table 1 Normalized mutual information of each sensor parameter

[0128]

[0129] After selecting 5 core nodes, the algorithm uses the Lagrange multiplier method to solve for the optimal relative humidity domain. The curing instructions for a 100mm thick slab are: initial stage (0-18h) relative humidity 92%, middle stage (18-36h) relative humidity 80%, and final stage (36-54h) relative humidity 60%. The measured relative humidity curves for each stage are shown below. Figure 3 As shown, Figure 3 The study demonstrates the change of the actual relative humidity in the curing chamber over time during the three stages of the gradient curing process, as well as the deviation between the algorithm's output target value and the measured value.

[0130] After curing, a full range of quality tests were conducted on the hardened boards. The test results are shown in Table 2.

[0131] Table 2 Quality Inspection Results of Hardened Boards

[0132]

[0133] like Figure 3 As shown, the relative humidity control accuracy at each stage of the gradient curing process is good, and the maximum deviation between the algorithm's output target value and the measured value is 1.3%, indicating that the closed-loop control of the multi-source heterogeneous feature correlation constraint evolution algorithm has high stability. The cross-section of the hardened board was cut and sampled, and the distribution of EPS lightweight aggregate particles in the cross-section was statistically analyzed using image analysis methods. The standard deviation of the cross-section particle density was 6.1%, lower than the set threshold of 8%, verifying the control effect of the dynamic fluid topology evolution compensation algorithm on uniform particle distribution. The histogram of the cross-section EPS lightweight aggregate particle density distribution is shown below. Figure 4 As shown, Figure 4 It reflects the distribution of EPS lightweight aggregate particle volume fraction in each equally divided interval along the height direction of the hardened board cross section, and is used to visually evaluate the uniformity of particle distribution.

[0134] The advancements of this invention compared to traditional methods are reflected in the following principles: Traditional processes rely on fixed-speed stirring and manual judgment, lacking real-time quantitative characterization of the particle distribution within the slurry. Buoyancy-driven particle segregation cannot be detected and corrected before molding. This invention establishes an indirect characterization mechanism for the particle distribution within the slurry through the measurable external signal of stirring motor torque fluctuations. This allows for the detection and compensation of particle segregation before molding is complete, fundamentally avoiding the problem of uneven cross-sectional properties caused by segregation. Traditional interface treatments rely on mechanical roughening, and interface bonding relies on physical interlocking. This invention introduces polar functional groups through plasma jets and combines this with the chemical bonding and penetration of polymer modifiers, elevating the interface bonding mechanism from a single physical interlocking to a synergistic effect of chemical bonding and physical anchoring. The improvement in interface peel strength exceeds the theoretical limit of purely physical treatment. Traditional curing relies on a fixed humidity regime, which cannot be adjusted in real time to meet the differentiated needs of boards of different thicknesses within a batch. This invention achieves adaptive and precise control of curing parameters through an AI-driven multi-source heterogeneous feature correlation constraint evolution algorithm, enabling a single curing chamber to simultaneously meet the optimal curing needs of boards of different thicknesses, thus solving the problem of quality dispersion within a batch from a fundamental perspective.

[0135] It should be noted that the variables involved in this invention are explained in detail in Table 3.

[0136] Table 3. Variable Explanation Table

[0137]

[0138] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An EPS composite lightweight structural panel, characterized in that, Composed of fiber cement board, interface transition layer and core material layer, the fiber cement board is located on both sides of EPS composite lightweight structural board, symmetrically arranged; the interface transition layer is located between the inner side of each fiber cement board and the core material layer, formed by spraying epoxy resin emulsion or polymer modified cement slurry onto the inner surface of the fiber cement board after low temperature and normal pressure plasma jet treatment, and bonding the core material layer and fiber cement board through the dual action of chemical bonding and physical anchoring; the core material layer is located between two fiber cement boards, and is formed by casting after uniformly mixing cement-based cementitious material, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent and multidimensional flexible fiber. The multidimensional flexible fiber forms a three-dimensional spatial network anchoring structure in the core material layer, and the core material layer contains closed microbubbles introduced by the air-entraining agent. The EPS composite lightweight structural board uses tongue and groove joints on all four sides.

2. The EPS composite lightweight structural panel according to claim 1, characterized in that, The fiber cement board is specifically formed by silicate cement, quartz sand and cellulose fibers through a flow casting or pressing method. It does not contain asbestos or halogen components, and the inner surface is treated with low-temperature and normal-pressure plasma jet to introduce polar functional groups.

3. The EPS composite lightweight structural panel according to claim 2, characterized in that, The EPS lightweight aggregate particles are specifically expanded polystyrene foam particles with a particle size range of 1–5 mm and a density range of 10–30. The volumetric doping level ranges from 15% to 40%; the modified cellulose ether is a product of hydrophobic modification of hydroxypropyl methylcellulose ether; the multidimensional flexible fiber is at least two of polypropylene fiber, basalt fiber, and polyvinyl alcohol fiber, with a doping level ranging from 0.8% to 1.5%. .

4. A method for producing EPS composite lightweight structural panels, characterized in that, Includes the following steps: Cement-based cementitious material, EPS lightweight aggregate particles, modified cellulose ether, air-entraining agent and multidimensional flexible fiber are added to the mixer according to the formula. The dynamic fluid topology evolution compensation algorithm is started to collect the torque fluctuation data of the mixing motor in real time. The mixing speed is adjusted according to the output result of the dynamic fluid topology evolution compensation algorithm to obtain a uniform slurry. The inner side of the fiber cement board is subjected to low-temperature atmospheric pressure plasma jet treatment. After the treatment is completed, an interface modifier is pre-sprayed on the inner side of the fiber cement board to form an interface transition layer. The uniform slurry is injected into a mold made of fiber cement board that has undergone low-temperature and normal-pressure plasma jet treatment. The high-frequency micro-amplitude vibration molding process is used, and the molding pressure is controlled by the mold prestress reverse compensation technology. The molded board is obtained by one-time composite molding. The multi-source heterogeneous feature correlation constraint evolution algorithm is activated to collect temperature sensor data, humidity sensor data and shrinkage sensor data in the curing environment in real time. Based on the output results of the multi-source heterogeneous feature correlation constraint evolution algorithm, the gradient curing process is executed. The relative humidity of the curing environment is precisely controlled in stages to complete the hardening of the molded board and obtain the hardened board. The hardened boards are tested sequentially for flatness deviation, diagonal difference, compressive strength, sound insulation and fire resistance limit. Hardened boards that pass all tests are then stacked and stored.

5. The method for producing EPS composite lightweight structural panels according to claim 4, characterized in that, The dynamic fluid topology evolution compensation algorithm is specifically based on the principle of the finite volume method. It establishes a coupled dynamic equation between the uniform slurry continuous medium and EPS lightweight aggregate particles, calculates the apparent viscosity of the flow field in real time, maps the density distribution probability map of EPS lightweight aggregate particles using coordinate transformation, and outputs the extrusion compensation amount at the outlet of the molding machine based on the particle density deviation.

6. The method for producing EPS composite lightweight structural panels according to claim 5, characterized in that, The torque fluctuation threshold range and the compression compensation adjustment coefficient in the dynamic fluid topology evolution compensation algorithm are specifically determined by systematically changing the water-cement ratio, EPS lightweight aggregate particle volume content, and stirring speed through no less than 30 sets of stirring molding experiments, with the goal of minimizing the standard deviation of cross-sectional EPS lightweight aggregate particle density through multiple iterative fitting.

7. The method for producing EPS composite lightweight structural panels according to claim 6, characterized in that, The aforementioned low-temperature atmospheric pressure plasma jet treatment specifically has an energy density range of 5–15. It is used to remove organic pollutants from the inner surface of fiber cement board and introduce polar functional groups; the interface modifier is epoxy resin emulsion or polymer modified cement slurry, and the thickness of the interface transition layer ranges from 0.3 to 0.8 mm.

8. The method for producing EPS composite lightweight structural panels according to claim 7, characterized in that, The high-frequency micro-amplitude vibration molding process specifically has a vibration frequency range of 50-100Hz and an amplitude range of 0.1-0.5mm. The vibration frequency range and amplitude range are determined by testing the content of closed micro-bubbles in the cross-section of the molded sheet and the uniformity of EPS lightweight aggregate particle distribution under different frequency and amplitude combinations.

9. The method for producing EPS composite lightweight structural panels according to claim 8, characterized in that, The aforementioned mold prestress reverse compensation technology specifically involves applying an elastic pre-deformation amount opposite to the direction of shrinkage deformation to the bottom plate and side plates of the mold in the mold design stage, based on the drying shrinkage law of inorganic materials.

10. The method for producing EPS composite lightweight structural panels according to claim 9, characterized in that, The gradient curing process specifically divides the hardening process into three stages: initial, intermediate, and final. The relative humidity is controlled within the range of 90% to 95% in the initial stage, 75% to 85% in the intermediate stage, and 55% to 65% in the final stage. The duration of each stage is adjusted according to the thickness of the molded board.

Images