Intelligent response type alkali-activated concrete and digital twin design and 4D printing method

Abstract

The application discloses an intelligent response type alkali-activated concrete and a digital twin design and 4D printing method, and belongs to the technical field of building materials and intelligent construction. The concrete has three-dimensional non-uniformly distributed functional materials inside, and the functional materials include intelligent response microcapsules with a four-layer structure of "core-shell-membrane-lubricating layer" and a hydrophobic agent. The application builds a new technology paradigm of "digital twin driven design, intelligent material creation, 4D printing precise construction, and whole-cycle intelligent management", realizes the precise distribution and dynamic regulation of the performance of the concrete material in the component according to needs, time and place, and synchronously solves key problems such as printing rheological regulation, microcapsule protection, interlayer penetration and sensor longevity, so that the durability, safety and service life of the concrete structure in a complex environment are greatly improved, and the application is suitable for long-life infrastructure engineering in extreme and harsh corrosive environments such as cross-sea bridges, undersea tunnels and saline-alkali land buildings.

Description

Technical Field

[0001] This invention relates to the field of building materials and intelligent construction technology, and in particular to an intelligent responsive alkali-activated concrete and a digital twin design and 4D printing method. Background Technology

[0002] Alkali-activated concrete (AAC), as a low-carbon cementitious material system, has attracted widespread attention due to its high strength, acid resistance, and low permeability. Its resistance to ionic attack can be further enhanced by incorporating hydrophobic agents (such as polysiloxanes). These polysiloxanes can form a dense hydrophobic film inside the concrete, effectively blocking Cl-. - SO4 2- The penetration of corrosive ions. However, existing technologies suffer from the following bottlenecks: 1) Inefficient material design methods: The dosage and distribution of hydrophobic agents (such as polysiloxane KH-560) and functional additives are mostly based on experience or homogenization assumptions. Typically, the dosage of polysiloxane hydrophobic agents is 0.5% to 2.0% of the mass of cementitious materials, and the dosage of functional additives is 1.0% to 3.0%. This makes it impossible to carry out precise spatial differentiation design based on the complex stress and environmental exposure differences inside the structure, resulting in material waste and insufficient local protection.

[0003] 2) Single and passive function: Traditional additives are mostly static and lack the ability to respond to corrosive environments (such as Cl). - The dynamic sensing and response capabilities for concentration, pH, and stress are insufficient to achieve "on-demand" protection during the corrosion process. When Cl in the environment... - When the concentration exceeds 0.05 mol / L and the pH is below 11.0, the protective effectiveness of traditional hydrophobic concrete will rapidly decline, and there is no active remediation mechanism.

[0004] 3) The preparation process is out of sync with the design: Even with an optimized design, the traditional mixing and pouring process is difficult to achieve a precise and gradient three-dimensional distribution of complex functional materials inside the concrete. The distribution deviation of functional materials can be more than 20%, which cannot meet the design expectations.

[0005] 4) Lack of full life cycle management: The lack of a data loop from design and construction to service maintenance makes predictive maintenance and performance optimization difficult. Existing alkali-activated concrete structures commonly exhibit surface powdering and steel corrosion after 10-15 years of service in marine environments, with annual maintenance costs accounting for 3%-5% of construction costs.

[0006] Although some material distribution studies based on finite element analysis have emerged in existing technologies, they generally remain at the level of single ion diffusion simulation and fail to integrate multi-physics coupling, artificial intelligence prediction and advanced digital manufacturing technologies, thus failing to achieve true collaborative intelligent design of "material-structure-environment". Summary of the Invention

[0007] The purpose of this invention is to provide a smart responsive alkali-activated concrete and a digital twin design and 4D printing method to solve the problems in the prior art.

[0008] To achieve the above objectives, the present invention provides a smart responsive alkali-activated concrete, wherein the concrete contains a three-dimensional non-uniformly distributed functional material, the functional material including smart responsive microcapsules and a hydrophobic agent; The intelligent responsive microcapsule has a four-layer structure of "core-shell-membrane-lubricating layer", specifically: The core contains a compound functional agent, which is at least two of the following: hydrophobic agent, corrosion inhibitor, microbial spores, and nanocrystalline nuclei; The shell is a smart response shell, capable of releasing functional agents from the core in a graded manner according to environmental stimuli; The outer compatibility membrane is a cement hydration product simulation layer modified with nano-SiO2, which is used to enhance the interfacial adhesion between the microcapsules and the alkali-activated matrix. The lubricating layer is a nanoscale self-lubricating coating that can generate a shear thinning effect under high shear rates during 4D printing, thus protecting the integrity of the microcapsules.

[0009] Preferably, in the compound functional agent, the hydrophobic agent is at least one of silane, siloxane, fluorocarbon resin and its modifiers; the corrosion inhibitor is a sodium molybdate / phytic acid composite corrosion inhibitor; the microbial spores are a mixture of Bacillus spores and nutrients, and the nanocrystal nuclei have a particle size of 5~20nm.

[0010] Preferably, the intelligent response shell is constructed using a layer-by-layer self-assembly technique, with a shell thickness of 10~30μm, a cross-linking degree of 60%~85%, and exhibits a graded response to the following stimuli: 1) It preferentially releases hydrophobic agents upon contact with moisture; 2) When the pH value drops below 10.5 or Cl... - Corrosion inhibitors are released when the concentration exceeds 0.05 mol / L; 3) When the local stress is ≥2MPa or cracks appear, microbial spores and nanocrystal nuclei are released.

[0011] Preferably, the nanoscale self-lubricating coating layer is nano-SiO2 or polyacrylamide derivative with surface grafted polyethylene glycol, with a mass fraction of 2%~5% and a thickness of 10~50nm. Under the high shear rate of 4D printing, the viscosity decreases by 60%~80%, and the retention rate after microcapsule printing is ≥98%.

[0012] Preferably, the concrete also contains shape memory polymer or temperature-sensitive + water-sensitive dual-response hydrogel, giving the concrete component a fourth-dimensional characteristic of shape recovery or performance evolution under environmental stimuli, with a shape recovery rate ≥90%, recovery stress sufficient to actively close microcracks with a width ≤0.3mm, and a response time ≤24h.

[0013] Preferably, the concrete is embedded with a distributed fiber optic sensor network and / or a wireless passive RFID sensor tag. The outer layer of the sensor is encapsulated with an alkali-resistant glass fiber reinforced polymer sleeve and is equipped with sacrificial anode protection technology to ensure the validity of 80 years of monitoring data.

[0014] Preferably, the intelligent responsive alkali-activated concrete comprises the following components: 1) Alkali-activated cementitious matrix materials: including alkali-activated cementitious materials, alkali activators, aggregates, and water; 2) Functional materials: including smart responsive microcapsules and hydrophobic agents; 3) 4D responsive materials: including shape memory polymer fibers and / or temperature-sensitive + water-sensitive dual-responsive hydrogels; 4) Printing process auxiliary materials: including thixotropic modifiers and fiber reinforcement materials; 5) Interlayer interface materials: including nano-silica sol and penetrating crystalline waterproofing agent; 6) Embedded sensing materials: including fiber optic sensors, RFID sensing tags, and corresponding encapsulation and protection materials.

[0015] This invention also provides a digital twin design and 4D printing method for the concrete, comprising the following steps: S1. Construct a high-fidelity digital twin model of the target concrete structure, extract holographic data including three-dimensional geometry, component topology, and design load, divide the target structure into non-uniform grids, and associate each grid cell with initial material properties and environmental boundary conditions. S2. Through multi-physics coupling and AI prediction models, simulate the performance evolution of the structure under service environment and output corrosion risk cloud map; S3. Using a deep reinforcement learning algorithm, with the dual goals of optimal life cycle cost and carbon footprint, a variety of preset functional materials are optimized and combined to generate a three-dimensional non-uniform material distribution digital map. S4. Based on the digital map of material distribution, concrete components are prepared using multi-material collaborative 4D printing technology, and in-situ monitoring and closed-loop control are achieved during the printing process.

[0016] Preferably, in S2, the multiphysics field includes ion migration, moisture diffusion, heat transfer, carbonization reaction, and mechanical stress damage; The AI ​​prediction model is a hybrid convolutional-recurrent neural network model. It uses transfer learning to correlate and label laboratory accelerated corrosion data with natural exposure data, with a correlation coefficient ≥0.9.

[0017] Preferably, in S4, the 4D printing technology adopts a multi-channel intelligent printing system, which includes an alkali-activated slurry delivery main channel, a hydrophobic agent precise atomization injection unit, an intelligent microcapsule quantitative doping device, a thixotropic agent dynamic addition channel, a directional fiber placement mechanism, an interlayer interface strengthening module, and an active triggering system. The interlayer interface strengthening module includes a laser processing unit and a penetrant spraying unit, which are used to construct a micro-nano rough interface on the surface of the formed layer and spray nano-silica sol or penetrating crystalline waterproofing agent to ensure that the interlayer bonding strength is ≥2.5MPa.

[0018] Preferably, during the 4D printing process, the modulus of the alkali activator and the distribution ratio of each functional group are dynamically adjusted according to the real-time feedback of the digital twin model to realize the gradient structure of material properties in three-dimensional space, and the length of the performance gradient transition zone is controlled at 50~100mm. During the printing process, the rheology of the extruded material, the uniformity of microcapsule distribution, and the quality of interlayer bonding are monitored in real time by near-field infrared spectroscopy and computer vision system, and the printing parameters are adjusted in real time by machine learning algorithm to achieve closed-loop control of the process.

[0019] Preferably, the prepared intelligent responsive alkali-activated concrete is applied to cross-sea bridges, subsea tunnels, and buildings in saline-alkali land under extreme corrosive environments.

[0020] Therefore, the intelligent responsive alkali-activated concrete and digital twin design and 4D printing method of the present invention have the following beneficial effects: (1) It has realized the leap from concrete to "homogeneous material" to "performance programmable intelligent component", integrated the digital design dimension of carbon footprint, upgraded the design basis from empirical formula to precise digital twin based on AI and multi-physics simulation, optimized both material usage and carbon emissions, improved the utilization rate of protective materials in high-risk areas, reduced material waste in low-risk areas, and reduced the carbon footprint throughout the entire cycle.

[0021] (2) The "core-shell-membrane-lubricating layer" quadruple structure and step-by-step release mechanism of the smart microcapsule realizes the "on-demand graded triggering" of protection and repair functions; the lubricating layer protects the integrity of the microcapsule during the 4D printing process with a retention rate of ≥98%; combined with the interlayer penetration enhancement layer technology, it blocks the erosion path from the source and reduces the chloride ion diffusion coefficient.

[0022] (3) By driving 4D printing through digital twin model, the material properties are precisely constructed in three-dimensional space with gradient, and the material distribution deviation is ≤3%; the modulus of thixotropic modulator and alkali activator is dynamically controlled, which solves the contradiction between "easy extrusion" and "fast hardening" of slurry.

[0023] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0024] Figure 1 This is a flowchart of the present invention; Figure 2 This is a diagram illustrating the structure and release mechanism of the smart microcapsule of the present invention; Figure 3 This is a schematic diagram of the three-dimensional material distribution design in Embodiment 1 of the present invention; Figure 4 This is a concentration distribution diagram of the microcapsules in Example 1 of the present invention; Figure 5 This is a spatial evolution diagram of carbon footprint based on material distribution in Embodiment 1 of the present invention; Figure 6 This is a diagram showing the internal structure of the multi-channel intelligent printing system of the present invention; Figure 7 This is a logic diagram of the digital twin data flow in Embodiment 1 of the present invention; Figure 8 This is a comparison curve of the weight loss rate of reinforcing bars in Embodiment 1 of the present invention; Figure label: 1. Lubricating layer; 2. Compatible film; 3. Shell; 4. Core. Detailed Implementation

[0025] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0027] This invention provides a smart responsive alkali-activated concrete, specifically comprising the following components: 1) Alkali-activated cementitious matrix materials: including alkali-activated cementitious materials, alkali activators, aggregates, and water; 2) Functional materials: including smart responsive microcapsules and hydrophobic agents; 3) 4D responsive materials: including shape memory polymer fibers and / or temperature-sensitive + water-sensitive dual-responsive hydrogels; 4) Printing process auxiliary materials: including thixotropic modifiers and fiber reinforcement materials; 5) Interlayer interface materials: including nano-silica sol and penetrating crystalline waterproofing agent; 6) Embedded sensing materials: including fiber optic sensors, RFID sensing tags, and corresponding encapsulation and protection materials.

[0028] Specifically, the concrete contains three-dimensionally non-uniformly distributed functional materials, including smart responsive microcapsules and hydrophobic agents; the hydrophobic agents are at least one of silanes, siloxanes, fluorocarbon resins and their modifiers.

[0029] like Figure 2 As shown, the intelligent responsive microcapsules possess a quadruple structure of "core-shell-membrane-lubricating layer," serving as the core functional carrier. The microcapsule particle size is controlled at 50–200 μm, and the bulk density is 1.2–1.4 g / cm³. 3 Encapsulation rate ≥85%, dispersion uniformity ≥90% in alkali-activated slurry; specifically: Core 4 contains a compound functional agent, with a total content of 40%~60% of the microcapsule mass. The compound functional agent is at least two of the following: a hydrophobic agent, a corrosion inhibitor, microbial spores, and nanocrystal nuclei; achieving a multi-stage release ratio. The hydrophobic agent is at least one of silane, siloxane, fluorocarbon resin, and their modified forms; the corrosion inhibitor is a sodium molybdate / phytic acid composite corrosion inhibitor; the microbial spores are at a concentration of Bacillus spores (10... 9 -10 10 A mixture of CFU / g and nutrients (glucose and peptone in a mass ratio of 2:1), with nano-CSH crystal nuclei having a particle size of 5~20nm.

[0030] Shell 3 is a smart responsive shell constructed using layer-by-layer self-assembly technology. It can release functional agents from the core in a graded manner according to environmental stimuli. The shell thickness is 10~30μm, and the degree of cross-linking is 60%~85%. The degree of cross-linking can be controlled by the polyelectrolyte concentration (0.5~2.0g / L), and it exhibits a graded response to the following stimuli: 1) Upon contact with moisture, it preferentially releases hydrophobic agents (first stage); 2) When the pH value drops below 10.5 or Cl... - When the concentration exceeds 0.05 mol / L, corrosion inhibitor is released (second stage); 3) When the local stress is ≥2MPa or cracks appear, microbial spores and nanocrystal nuclei are released (Level 3).

[0031] The outer compatibility membrane 2 is a cement hydration product simulation layer modified with nano-SiO2. The SiO2 particle size is 20~50nm, the dosage is 15%~25% of the membrane mass, and the membrane thickness is 5~10μm. It ensures that the interfacial bonding strength between the microcapsules and the alkali-activated slurry is ≥3.5MPa, and the microcapsule retention rate is ≥90% after curing in the slurry for 28 days. Lubricating layer 1 is a nanoscale self-lubricating coating layer with a thickness of 10-50 nm. It uses a shear-thinning fluid, such as nano-SiO2 grafted with polyethylene glycol or a polyacrylamide derivative, with a mass fraction of 2%-5%. Under the high shear rate of 4D printing, it produces a shear-thinning effect, reducing viscosity by 60%-80%, forming a fluid lubrication protective layer, which improves the retention rate of microcapsules after printing to over 98%. At the same time, the microcapsules adopt a multi-stage release mechanism. The first stage releases a hydrophobic agent to block moisture and corrosive ions, the second stage releases a corrosion inhibitor to protect the steel bars from corrosion, and the third stage releases microorganisms and nanocrystal nuclei to repair cracks, forming a stepped protection system.

[0032] The concrete also contains shape memory polymers or thermo- and water-sensitive dual-response hydrogels. By incorporating 1.5% to 3.0% (mass fraction) of plasma-modified shape memory polymers (SMP, melting point 40 to 60°C) or thermo- and water-sensitive dual-response hydrogels (water absorption expansion rate ≥50%) into the alkali-activated matrix, the concrete component possesses a fourth-dimensional characteristic of shape recovery or performance evolution under environmental stimuli. The shape recovery rate is ≥90%, the recovery stress is sufficient to actively close microcracks with a width ≤0.3 mm, and the response time is ≤24 h. The active groups formed on the surface of the SMP fibers through etching treatment form chemical bonds with the alkali-activated concrete (AAC) matrix.

[0033] The concrete structure is embedded with a distributed fiber optic sensor network and / or wireless passive RFID sensor tags (reading distance 5~10m, storage capacity ≥1024 bytes). The sensor deployment density is no less than 5 per cubic meter for long-term monitoring of strain, temperature, and humidity. To ensure the long lifespan of the sensors, the outer layer of the fiber optic sensors is encapsulated with an alkali-resistant glass fiber reinforced polymer (BFRP) sleeve with a thickness of 2~3mm and an alkali resistance rating ≥pH14. Sacrificial anode protection technology is also employed, with zinc alloy sacrificial anodes (mass fraction 0.5%~1.0%) deployed around the sensors to slow down the alkali corrosion aging of the sensor encapsulation layer. The sensor attenuation curve is recorded based on blockchain technology, and data compensation is performed through AI algorithms to ensure the validity of 80 years of monitoring data.

[0034] This invention also provides a digital twin design and 4D printing method for the concrete, such as... Figure 1 As shown, it includes the following steps: S1. Construct a high-fidelity digital twin model of the target concrete structure, extract holographic data including 3D geometry, topological relationships, and design loads, and perform non-uniform meshing on the target structure. Automatically refine the mesh in high-risk areas (such as corners and splash zones) with a mesh size of 3-5mm; and in medium- and low-risk areas with a mesh size of 15-25mm. Generate a set of elements containing spatial coordinates, volume, and exposed surface identifiers. The number of mesh elements in a single pier model is 800,000 to 1,200,000.

[0035] Each grid cell is associated with initial material properties and environmental boundary conditions. The initial material is an alkali-activated cementitious matrix, and the material properties are compressive strength, flexural strength, and chloride ion diffusion coefficient. The environmental boundary conditions are obtained from a database containing chloride ions from different sea areas. - Dynamic environmental parameters such as concentration, pH value, and temperature.

[0036] S2. By using multiphysics coupling and AI prediction models, the performance evolution of the structure under service environment is simulated, and a corrosion risk cloud map is output; specifically, the multiphysics field includes ion migration (Cl... - SO4 2- ( ), moisture diffusion, heat transfer, carbonization reaction and mechanical stress damage; the coupling error of the multiphysics coupling model is ≤5%, of which the diffusion coefficient correction coefficient of the ion migration model is 0.85~0.95.

[0037] The AI ​​prediction model is a hybrid convolutional-recurrent neural network (CNN-LSTM) model, trained with massive amounts of historical corrosion data (1000+ sets of marine environment concrete corrosion data and 500+ sets of laboratory accelerated corrosion data) to predict the spatiotemporal evolution of key corrosion factor concentrations, pH values, and rebar depassivation risk within each unit. The model training accuracy is ≥92%, and the prediction error is ≤8%. Through transfer learning, laboratory accelerated test data (7-day dry-wet cycle, 3.5% NaCl concentration) are correlated and calibrated with actual natural exposure data, with a correlation coefficient ≥0.9 between the accelerated test and natural exposure data. This yields the NaCl concentration on the bridge pier surface for the next 50 years. - Concentration distribution, carbonization depth, and stress concentration zone evolution cloud maps were used to identify the tidal fluctuation zone and the back wave face corner as high-risk corrosion areas.

[0038] Monte Carlo simulations were performed to quantify material parameters (hydrophobic agent content fluctuation ±0.2%, microcapsule particle size fluctuation ±5μm, thixotropic modifier content fluctuation ±0.1%) and environmental loads (Cl). - The uncertainty of concentration fluctuation (±0.01mol / L) is used to output a corrosion development probability cloud map characterized by reliability indicators (such as failure probability ≤5%).

[0039] S3. Using a deep reinforcement learning algorithm, with the dual goals of optimizing the whole life cycle cost and carbon footprint, a variety of preset functional materials are optimized and combined to generate a three-dimensional non-uniform material distribution digital map.

[0040] Before performing optimization combinations, it is necessary to first establish a smart response material library, which includes: Various types of smart responsive microcapsules, and their encapsulation efficiency and release kinetic parameters; Various hydrophobic agents (such as polysiloxane KH-560, silanes), and their hydrophobic effects and cost data; Various 4D responsive materials (shape memory polymers, dual-response hydrogels), and their response temperatures and recovery rates; Carbon emission coefficients for each material (calculated based on the LCA method).

[0041] Based on this material library, the deep reinforcement learning algorithm optimizes the material combinations of each grid cell and finally outputs a digital map of material distribution.

[0042] Select functional materials and define an objective function: minimize "material cost + expected maintenance cost + failure risk cost + total lifecycle carbon emissions", where material cost accounts for 60%~70%, expected maintenance cost accounts for 20%~25%, failure risk cost accounts for 5%~10%, and carbon emissions account for 3%~5%. Embed a real-time carbon emission calculation module into the digital twin model. Based on the life cycle assessment (LCA) method, quantify the carbon emission coefficients of each material (cementing materials, hydrophobic agents, microcapsules, thixotropic modifiers, etc.), calculate the carbon footprint of different material distribution schemes in real time, and achieve optimal carbon footprint during the design phase.

[0043] The constraints include: the probability of steel bar debuffing in each unit within the specified service life (80 years) is less than 5%; the self-healing efficiency of 0.3mm cracks after 14 days is ≥85%; the hydrophobic agent dosage is 0.3%~3.0% of the cementitious material mass; the microcapsule dosage is 1.0%~4.0% of the cementitious material mass; the printing speed is ≥50mm / s and the forming accuracy is ±1mm; the interlayer bond strength is ≥2.5MPa; and the carbon footprint throughout the entire cycle is reduced by more than 15% compared with traditional alkali-activated concrete.

[0044] Deep reinforcement learning algorithms (such as Proximal Policy Optimization, PPO) are used for optimization, with a learning rate of 0.001-0.003, 20,000-30,000 iterations, and a convergence error ≤3%. Based on the corrosion evolution path predicted by the digital twin, the carbon footprint calculation results, and the printing process requirements, the algorithm dynamically adjusts the type, dosage, hydrophobic agent concentration, and thixotropic modifier dosage of the smart microcapsules within each 3D grid cell. Leveraging the highly efficient hydrophobic properties of the hydrophobic agent and the synergistic effect of the microcapsule's stepped protection and thixotropic regulation, the algorithm ultimately outputs a 3D, non-uniform "material distribution digital map" with a resolution ≤5mm and a material distribution deviation ≤3%.

[0045] S4. Based on the digital map of material distribution, concrete components are prepared using multi-material collaborative 4D printing technology, and in-situ monitoring and closed-loop control are achieved during the printing process.

[0046] Specifically, the design involves a multi-channel intelligent printing system with a rated power of 15-20kW and an effective printing space of 5m×3m×8m. The system is equipped with a pose closed-loop compensation module based on a laser tracker to ensure end-effector positioning accuracy of ±0.2mm under large-scale cantilever conditions.

[0047] like Figure 6 As shown, the system (printhead) includes an alkaline-activated slurry delivery main channel (inner diameter 50~80mm, pressure 0.3~0.8MPa), a hydrophobic agent precision atomization injection unit (atomized particle size 10~50μm, accuracy ±0.01%), an intelligent microcapsule quantitative dosing device (delivery rate 0.1~1.0kg / min, accuracy ±2%), a thixotropic agent dynamic addition channel (accuracy ±0.01%), and a directional fiber placement mechanism (fiber diameter 10~20μm, length 6~12mm, placement density 0~2kg / m²). 3 Interlayer interface strengthening module and active triggering system (including micron-level heating wire embedding device or chemical trigger injector).

[0048] The interlayer interface strengthening module includes a laser processing unit and a penetrant spraying unit (spraying accuracy ±0.1mm, spraying amount 0.1~0.3kg / m). 2 This is used to construct a micro-nano rough interface on the surface of the formed layer and spray nano-silica sol or penetrating crystalline waterproofing agent to ensure interlayer bonding strength ≥2.5MPa.

[0049] The system parses the "material distribution digital map" obtained in S3 into a 4D printing path containing spatiotemporal response commands, with a path planning efficiency of ≥10,000 grid cells / minute. During the printing process, the system uses a laser displacement meter and a thermal infrared imager integrated at the execution end to collect real-time online data on the six-DOF pose of the print head, instantaneous interlayer settlement height, and temperature field distribution in the forming area. The collected multidimensional parameters are preprocessed by edge computing nodes and then synchronized to a high-fidelity digital twin model. This model uses a real-time multiphysics coupling algorithm to evaluate the current forming quality online and predict the subsequent 4D temporal evolution trend of the component. Based on the real-time feedback and predicted trend of the digital twin model, the system dynamically adjusts the modulus of the alkali activator, the dosage of the thixotropic modifier, and the extrusion rate and ratio of each functional material through a dynamic mixing chamber located at the end of the print head. This achieves a gradient construction of material properties in three-dimensional space, with the length of the performance gradient transition zone controlled within 50~100mm to ensure no stress abrupt changes at the heterogeneous interface.

[0050] The modulus of the alkali activator is adjusted within the range of SiO2 / Na2O ratio of 2.0 to 2.8; the dosage of the thixotropic modifier is 0.1% to 0.5%; and the extrusion rates and proportions of each functional material are as follows: main slurry extrusion rate 5 to 20 L / min, hydrophobic agent 0.01 to 0.06 L / min, and microcapsules 0.05 to 0.2 kg / min.

[0051] Simultaneously, for high-risk exposure surfaces identified by the digital twin model, the system implements an interlayer synergistic enhancement process: First, a micro-nano rough interface (Ra value 3~10μm) is constructed on the surface of the formed layer using a laser processing unit; subsequently, a high-concentration nano-silica sol or penetrating crystalline waterproofing agent is uniformly sprayed using a spraying unit. The system automatically identifies the edge areas of the component and performs a 50% spraying increment compensation, strengthening the interlayer seal through a dual mechanism of physical anchoring and chemical bonding, blocking the chloride ion penetration path, and ensuring the structural integrity during the 4D evolution process.

[0052] During the printing process, near-field infrared spectroscopy (detection range 4000-4000 cm⁻¹) is used. -1 The system uses a combination of detection accuracy ±0.1% and a computer vision system (frame rate 30~60fps, recognition accuracy ±0.1mm) to monitor the rheological properties of extruded materials in real time (yield stress 200-500Pa, plastic viscosity 100-300Pa). The system ensures uniform microcapsule distribution and interlayer bonding quality (interlayer bonding strength ≥ 2.5 MPa), and uses machine learning algorithms to adjust printing parameters in real time to achieve closed-loop process control, resulting in a printing quality pass rate ≥ 98%.

[0053] In addition, based on blockchain technology (using a consortium blockchain architecture with 5-10 nodes), an immutable "digital passport" is established for each concrete component produced using the above method. This passport records the design parameters (material distribution, grid division data, carbon footprint data), material sources (batch of cementitious materials, microcapsule manufacturer, type of hydrophobic agent), printing process data (printing speed, pressure, amount of each material, alkali activator modulus adjustment record), sensor attenuation curves, and data compensation records. The data storage capacity is ≥10GB, and the access latency is ≤50ms.

[0054] During its service life, data is continuously collected through built-in sensors and external detection methods to update the digital twin of components, enabling real-time health assessment and predictive maintenance decision support.

[0055] The above technical solution will be further explained below with reference to specific embodiments.

[0056] Example 1 This embodiment takes the pier revetment layer in the splash zone of a cross-sea bridge as an example. This cross-sea bridge is located in a temperate maritime climate zone with seawater Cl... - With a concentration of 0.06 mol / L and an average annual temperature of 18℃, the bridge piers in the splash zone withstand a wave force of 1.0 kN / m. 2 The design protective layer is 300mm thick, with a target service life of 80 years. It requires that the probability of steel bar descaling be less than 5% within 30 years, the self-healing efficiency of 0.3mm cracks be ≥85% after 14 days, and the interlayer bond strength be ≥2.5MPa.

[0057] Digital twin data flow logic as follows Figure 7 As shown, the specific steps for creating intelligent alkali-activated concrete using digital twin design and 4D printing methods are as follows: S1. Obtain the bridge pier BIM model (based on the LOD400 precision design model with reserved LOD500 operation and maintenance interface), extract the three-dimensional geometric dimensions (diameter 2.5m, height 15m), topological relationships, and load data (wave force 1.0kN / m). 2 Self-weight load 25kN / m 3 The data acquisition error is controlled within ±0.5mm.

[0058] The non-uniform grid was divided using the leading edge method, generating approximately 1 million grid cells. The grid size was 5 mm in the splash zone (within ±2 m elevation), 10 mm in the tidal zone (±2~5 m elevation), and 20 mm in the internal region (elevation >5 m and < -2 m). Each cell was associated with spatial coordinates, volume, and exposed surface identification.

[0059] Enter 50 years of local marine environmental data: Cl -Concentration 0.04~0.08mol / L (fluctuation ±0.01mol / L), pH 7.8~8.5, temperature 5~32℃, wave force 0.8~1.2kN / m 2 The initial performance parameters of the alkali-activated gel matrix were correlated (28-day compressive strength 62 MPa, chloride ion diffusion coefficient 0.9 × 10⁻⁶). -12 m 2 The carbon emission coefficients of each material (metakaolin 0.15kgCO2 / kg, slag 0.05kgCO2 / kg, polysiloxane KH-560 0.8kgCO2 / kg, nano clay 0.3kgCO2 / kg) are also included.

[0060] S2. A trained CNN-LSTM hybrid model (93.5% accuracy, 7.2% prediction error) was used for simulation. Transfer learning was applied to correlate accelerated laboratory experiments (3.5% NaCl solution, 7-day wet-dry cycle, 25℃) with natural exposure data (correlation coefficient 0.92) to obtain the Cl- concentration on the bridge pier surface over the next 50 years. - Concentration distribution, carbonization depth, and stress concentration zone evolution cloud maps were used to identify the tidal fluctuation zone and the back wave face corner as high-risk corrosion areas.

[0061] 10,000 Monte Carlo simulations were performed to quantify the fluctuations in hydrophobic agent content (±0.2%), microcapsule particle size (±5 μm), nano-clay content (±0.1%), and Cl... - The uncertainty of concentration fluctuation ±0.01mol / L is used to output a failure probability cloud map. The initial failure probability in the high-risk area is 12.3%, which needs to be reduced to below 5% through material optimization.

[0062] S3, Intelligent Material Distribution Optimization Design: Select the following functional materials: pH / Cl -Dual-responsive microcapsules (particle size 80~150μm, encapsulation efficiency 86%, shell thickness 20μm, crosslinking degree 75%, outer layer coated with a cement hydration product simulation layer modified with nano-SiO2, lubrication layer is surface-grafted polyethylene glycol nano-SiO2), stress-responsive microcapsules (particle size 100~200μm, encapsulation efficiency 88%, shell thickness 25μm, fracture stress 2MPa, also equipped with an outer compatibility membrane and lubrication layer), hydrophobic agent is hydrophobic polysiloxane (KH-560, purity ≥98%, viscosity 20). The materials include: -30 mPa·s, alkali-activated cementitious material (metakaolin: slag = 70:30, silica-alumina ratio 3.2:1), thixotropic modifier (nanoclay, particle size 50~100nm), and interlayer spraying material (high-concentration nano-silica sol, solid content 30%). Among them, polysiloxane KH-560 is well compatible with alkali-activated slurry, and the nano-clay content is 0.2%~0.4% to ensure that the slurry rheology meets the standards. The 4D responsive material is a temperature-sensitive + water-sensitive dual-responsive hydrogel (volume phase change temperature 35℃, water absorption expansion rate 60%).

[0063] 2) Set the objective function: minimize the total life cycle cost and carbon emissions, where material costs account for 65%, maintenance costs account for 22%, failure risk costs account for 13%, and carbon emissions account for 4%; Constraints: Probability of steel bar debuffing within 30 years <5%, self-healing efficiency of 0.3mm cracks ≥85% after 14 days, hydrophobic agent dosage 0.3%~3.0%, microcapsule dosage 1.0%~4.0%, printing speed ≥50mm / s, forming accuracy ±1mm, interlayer bond strength ≥2.5MPa, and carbon footprint reduction of more than 18%.

[0064] The PPO algorithm was used for 25,000 iterations of optimization (learning rate 0.002, convergence error 2.8%), outputting the optimal material distribution schemes for high-risk areas (tidal fluctuation zones and corners), medium-risk areas (tidal range zones, excluding corners), and low-risk areas (internal and non-tidal range zones). Figures 3-5 As shown in Table 1.

[0065] Table 1: Optimal Material Distribution Scheme

[0066] The material distribution deviation is 2.5%, which meets the design requirements.

[0067] S4, 4D printing preparation: Printing material ratio (by mass): 100 parts geopolymer cementitious material (70% metakaolin + 30% slag); water-cement ratio 0.32; 25 parts alkali activator (water glass: sodium hydroxide = 3:1, modulus dynamically adjusted according to region 2.2~2.6); 150 parts quartz sand (particle size 0.1~2.0mm); 2.0 parts thermosensitive + water-sensitive dual-response hydrogel (volume phase change temperature 35℃, water absorption expansion rate 60%); 1.5 parts steel fiber (diameter 15μm, length 8mm); 0.2%~0.4% nano clay (thixotropic modifier) ​​(adjusted according to region); 1 part water (to adjust rheology).

[0068] The printing path is programmed based on the material distribution map, with a path resolution of 5 mm, a print head movement speed of 60 mm / s, a main slurry delivery pressure of 0.5 MPa, and an extrusion rate of 12 L / min.

[0069] Printing parameters for each area are dynamically adjusted according to the optimization scheme, as follows: When printing in high-risk areas, the hydrophobic agent atomization injection rate is 0.06 L / min, the dual-response microcapsule delivery rate is 0.175 kg / min, the stress-response microcapsule delivery rate is 0.025 kg / min, the nano-clay addition rate is 0.004 kg / min, and the alkali activator modulus is adjusted to 2.6.

[0070] The hydrophobic agent injection rate in the medium-risk area was 0.03 L / min, the total delivery rate of the microcapsules was 0.1 kg / min, the nano-clay addition rate was 0.003 kg / min, and the modulus of the alkali activator was adjusted to 2.4.

[0071] The injection rate of hydrophobic agent in low-risk areas was 0.01 L / min, the total delivery rate of microcapsules was 0.05 kg / min, the addition rate of nano-clay was 0.002 kg / min, and the modulus of alkali activator was adjusted to 2.2.

[0072] After each layer is printed, a nano-silica sol interlayer penetration reinforcement layer is sprayed onto high-risk areas. Sandblasting is then used to achieve an interlayer interface roughness Ra of 5 μm, with a spraying amount of 0.25 kg / m². 2 The coating amount in the edge areas was increased to 0.375 kg / m². 2 .

[0073] During the printing process, near-field infrared spectroscopy and computer vision systems are used to monitor rheology (yield stress 380 Pa, plastic viscosity 220 Pa·s) and interlayer bonding quality (interlayer bonding strength 2.8 MPa) in real time. Printing parameters are adjusted through machine learning algorithms to ensure a print quality pass rate of 98.5%.

[0074] In addition, during its service life, data is collected through built-in sensors and external detection methods such as drones and robots to continuously update its digital twin, enabling real-time health assessment and predictive maintenance decision support, with a maintenance early warning accuracy rate of ≥90%.

[0075] The printed protective layer specimens were subjected to performance tests. The results showed that the 28-day compressive strength was 67 MPa, the flexural strength was 7.5 MPa, and the chloride ion diffusion coefficient was 0.4 × 10⁻⁶. -12 m 2 / s, contact angle of 112°, 14d self-healing efficiency of 0.3mm crack of 88%, interlayer bond strength of 2.8MPa.

[0076] After immersion in 3.5% NaCl solution for 500 days, the corrosion current density of the steel reinforcement was 0.04 μA / cm. 2 The total carbon footprint is reduced by 19.2% compared to traditional methods. All indicators meet the design requirements.

[0077] Comparative Example 1 This comparative example uses traditional hydrophobic alkali-activated concrete without the addition of smart microcapsules and employs conventional pouring techniques.

[0078] The concrete from Example 1 and Comparative Example 1 were immersed in a 3.5% NaCl solution for 500 days, and the corrosion of the reinforcing steel was analyzed. Figure 8 As shown, the results indicate that hydrophobic agents can delay corrosion, but their protective capabilities weaken in the later stages. In this Example 1, the synergistic effect of hydrophobic barrier, corrosion inhibitor chemical protection, and microbial mineralization repair, along with the process guarantee of precise 4D printing distribution and lubrication layer protection of microcapsule integrity, significantly improves the structural density and resistance to chloride ion penetration, demonstrating the long-term durability of the technical solution of this invention in extreme corrosive environments.

[0079] Example 2 This embodiment uses an undersea tunnel as the application scenario. The undersea tunnel is located in a sea area with high chloride ion concentrations. - Concentration 0.08 mol / L, water pressure 0.6 MPa, designed lining thickness 400 mm, service life target 80 years.

[0080] Using the same method as in Example 1, the intelligent concrete for the lining of the submarine tunnel was prepared by focusing on water pressure seepage and stress coupling in the digital twin model.

[0081] Similarly, the tunnel BIM model was obtained, and its geometric dimensions (inner diameter 10m, outer diameter 10.8m, length 100m segments) and load data (water pressure 0.6MPa, surrounding rock pressure, self-weight) were extracted. Non-uniform grid division was adopted, focusing on the high water pressure areas at the bottom of the tunnel and the lower part of the sidewalls, with the grid size refined to 5mm, and the grid size of other areas 15~20mm.

[0082] Input environment data: Cl - The concentration was 0.08 mol / L (fluctuation ±0.01 mol / L), pH 7.5–8.2, temperature 8–15℃, and water pressure 0.5–0.7 MPa. A multiphysics coupling model was used to simulate the coupling effect of water pressure permeation and stress.

[0083] By adding a water pressure permeation module to the digital twin model, AI predictions showed that the bottom of the tunnel and the lower part of the sidewalls were the most severe areas. After optimization, the dosage of stress-responsive microcapsules was increased to 1.0% in the high water pressure area, and an active triggering electrothermal wire network (50 μm diameter, 10 mm spacing) and an additional shape memory polymer fiber reinforcement network were set in this area.

[0084] During 4D printing, oriented steel fibers are simultaneously laid in a high water pressure area (layout density 1.8 kg / m²). 3 And spray a waterproof interlayer penetration reinforcement layer (a mixture of nano-silica sol and penetrating crystalline waterproofing agent, with a spraying amount of 0.3 kg / m²) onto the outer surface of the lining. 2 ).

[0085] The water permeability coefficient of the molded product is ≤1.0×10⁻⁶ after testing. -13 With a speed of m / s, it exhibits no leakage under a water pressure of 0.8MPa and a 28-day compressive strength of 72MPa, meeting the stringent service environment requirements of submarine tunnels.

[0086] Therefore, this invention provides a smart responsive alkali-activated concrete and a digital twin design and 4D printing method. The resulting concrete product has a non-uniform distribution of functional materials customized according to the digital twin design in three-dimensional space. It internally embeds a sensor network and smart responsive microcapsules encapsulated with alkali resistance and protected by sacrificial anodes, possessing environmental perception, adaptive stepped protection, and damage self-healing capabilities. Key performance indicators of the product: 28-day compressive strength ≥ 65 MPa, flexural strength ≥ 7 MPa, chloride ion diffusion coefficient ≤ 0.5 × 10⁻⁶. -12 m 2 / s, contact angle ≥110°, self-healing efficiency of 0.3mm cracks ≥85% after 14 days, interlayer bond strength ≥2.5MPa, and corrosion current density of steel bars ≤0.05μA / cm² after immersion in 3.5% NaCl solution for 500 days. 2 The carbon footprint over the entire lifecycle is reduced by more than 15% compared to traditional alkali-activated concrete.

[0087] This invention is designed with the dual goals of optimizing total life-cycle cost and carbon footprint. While initial material costs are manageable, overall costs are reduced by over 60% through significantly reduced maintenance frequency, extended lifespan, and lower carbon emissions. Simultaneously, the extended sensor lifespan solution ensures 80 years of monitoring effectiveness, and the intelligent management system reduces structural failure risk by over 80%, greatly improving operational safety and the scientific basis of decision-making. Furthermore, it deeply integrates multiple technologies, constructing a complete innovative system covering the entire chain from design to materials to construction to operation and maintenance. It specifically addresses key technical bottlenecks such as printing rheology control, microencapsulation protection, interlayer permeation, and sensor lifespan extension, improving the synergistic efficiency of various technologies by over 30%, providing a new technological path for the construction of infrastructure in extreme environments.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A smart responsive alkali-activated concrete, characterized in that: The concrete contains a three-dimensional non-uniformly distributed functional material, which includes smart responsive microcapsules and hydrophobic agents. The intelligent responsive microcapsule has a four-layer structure of "core-shell-membrane-lubricating layer", specifically: The core contains a compound functional agent, which is at least two of the following: hydrophobic agent, corrosion inhibitor, microbial spores, and nanocrystalline nuclei; The shell is a smart response shell, capable of releasing functional agents from the core in a graded manner according to environmental stimuli; The outer compatibility membrane is a cement hydration product simulation layer modified with nano-SiO2, which is used to enhance the interfacial adhesion between the microcapsules and the alkali-activated matrix. The lubricating layer is a nanoscale self-lubricating coating that can generate a shear thinning effect under high shear rates during 4D printing, thus protecting the integrity of the microcapsules.

2. The intelligent responsive alkali-activated concrete according to claim 1, characterized in that: In the compound functional agent, the hydrophobic agent is at least one of silane, siloxane, fluorocarbon resin and its modifiers; the corrosion inhibitor is a sodium molybdate / phytic acid composite corrosion inhibitor; the microbial spores are a mixture of Bacillus spores and nutrients, and the nanocrystal nuclei have a particle size of 5~20nm.

3. The intelligent responsive alkali-activated concrete according to claim 1, characterized in that: The intelligent response shell is constructed using a layer-by-layer self-assembly technique, with a shell thickness of 10–30 μm and a cross-linking degree of 60%–85%, and exhibits a graded response to the following stimuli: 1) It preferentially releases hydrophobic agents upon contact with moisture; 2) When the pH value drops below 10.5 or Cl... - Corrosion inhibitors are released when the concentration exceeds 0.05 mol / L; 3) When the local stress is ≥2MPa or cracks appear, microbial spores and nanocrystal nuclei are released.

4. The intelligent responsive alkali-activated concrete according to claim 1, characterized in that: The nanoscale self-lubricating coating is a nano-SiO2 or polyacrylamide derivative grafted with polyethylene glycol on its surface, with a mass fraction of 2%~5% and a thickness of 10~50nm. Under the high shear rate of 4D printing, the viscosity decreases by 60%~80%, and the retention rate after microcapsule printing is ≥98%.

5. The intelligent responsive alkali-activated concrete according to claim 1, characterized in that: The concrete also contains shape memory polymers or temperature- and water-sensitive dual-response hydrogels, giving the concrete components a fourth dimension of characteristics, such as shape recovery or performance evolution under environmental stimuli.

6. The intelligent responsive alkali-activated concrete according to claim 1, characterized in that: The concrete is embedded with a distributed optical fiber sensor network and / or a wireless passive RFID sensor tag, and the sensor is encapsulated in an alkali-resistant glass fiber reinforced polymer sleeve.

7. A method for digital twin design and 4D printing of concrete as described in any one of claims 1-6, characterized in that, Includes the following steps: S1. Construct a high-fidelity digital twin model of the target concrete structure, divide the target structure into non-uniform meshes, and associate each mesh cell with initial material properties and environmental boundary conditions; S2. Through multi-physics coupling and AI prediction models, simulate the performance evolution of the structure under service environment and output corrosion risk cloud map; S3. Using a deep reinforcement learning algorithm, with the dual goals of optimal life cycle cost and carbon footprint, a variety of preset functional materials are optimized and combined to generate a three-dimensional non-uniform material distribution digital map. S4. Based on the digital map of material distribution, concrete components are prepared using multi-material collaborative 4D printing technology, and in-situ monitoring and closed-loop control are achieved during the printing process.

8. The method according to claim 7, characterized in that: In S2, the multiphysics field includes ion migration, moisture diffusion, heat transfer, carbonization reaction, and mechanical stress damage. The AI ​​prediction model is a hybrid convolutional-recurrent neural network model. It uses transfer learning to correlate and label laboratory accelerated corrosion data with natural exposure data, with a correlation coefficient ≥0.

9.

9. The method according to claim 7, characterized in that: In S4, the 4D printing technology adopts a multi-channel intelligent printing system, which includes an alkali-activated slurry delivery main channel, a hydrophobic agent precise atomization injection unit, an intelligent microcapsule quantitative doping device, a thixotropic agent dynamic addition channel, a directional fiber placement mechanism, an interlayer interface strengthening module, and an active triggering system. The interlayer interface strengthening module includes a laser processing unit and a penetrant spraying unit, which are used to construct a micro-nano rough interface on the surface of the formed layer and spray nano-silica sol or penetrating crystalline waterproofing agent to ensure that the interlayer bonding strength is ≥2.5MPa.

10. The method according to claim 7, characterized in that: During the 4D printing process, the modulus of the alkali activator and the distribution ratio of each functional group are dynamically adjusted according to the real-time feedback of the digital twin model to realize the gradient structure of material properties in three-dimensional space, and the length of the performance gradient transition zone is controlled at 50~100mm. During the printing process, the rheology of the extruded material, the uniformity of microcapsule distribution, and the quality of interlayer bonding are monitored in real time by near-field infrared spectroscopy and computer vision system, and the printing parameters are adjusted in real time by machine learning algorithm to achieve closed-loop control of the process.

Images