Preparation method of high-insulation flame-retardant spraying polyurea elastomer material

Abstract

The invention provides a preparation method of a high-insulation flame-retardant spraying polyurea elastomer material, and belongs to the technical field of polyurea elastomer materials, the preparation method comprises the following steps: firstly, preparing an isocyanate component and a resin component, the isocyanate component comprises a rigid monomer and a flexible prepolymer, and the resin component comprises a rigid monomer and a flexible monomer; and calculating the optimal composition proportion meeting the requirements of target molecular weight, crosslinking density and mechanical property by utilizing a genetic algorithm. Mixing the two components in proportion, adding an auxiliary agent, and carrying out stirring reaction to form an intermediate solution; and carrying out reduced-pressure heating treatment on the solution to prolong the reaction time. And pouring or spraying the treated solution on a sampling plate, and curing to form a cross-linked polyurea elastomer coating. And finally, carrying out physical and chemical property tests on the cured material, wherein the physical and chemical properties comprise tensile strength, elongation at break, insulating property, flame retardant property and the like. A genetic algorithm is adopted to optimize an algorithm auxiliary formula design, and the research and development period is shortened.

Description

Technical Field

[0001] The invention belongs to the technical field of polyurea elastomer materials, and in particular relates to a method for preparing a highly insulating and flame-retardant spray polyurea elastomer material. Background Art

[0002] Polyurea elastomer is an important type of thermosetting elastomer material, which is mainly prepared by addition polymerization of isocyanate prepolymer and amino resin, and has the advantages of excellent mechanical properties, excellent wear resistance and good chemical corrosion resistance. Polyurea elastomer is an elastomer material generated by addition polymerization of isocyanate components and amino compounds, which has the advantages of zero VOC, good mechanical properties, excellent wear resistance and good chemical corrosion resistance. By adjusting the type and composition ratio of raw monomers, the mechanical properties, hardness, thermal stability and chemical stability of polyurea elastomer can be effectively regulated. Commonly used rigid isocyanate monomers include 4,4'-diphenylmethane diisocyanate (MDI), 1,5-naphthalene diisocyanate (NDI), etc.; flexible isocyanate prepolymers mainly include polyester and polyether isocyanate prepolymers. Rigid polyol monomers are generally trihydroxy or higher hydroxyl aromatic compounds, such as 1,1,1-tris(4-hydroxyphenyl)ethane (THPE), 1,3,5-tris(2-hydroxyphenyl)benzene, rigid amine chain extenders diethyltoluenediamine, dimethylthiotoluenediamine, etc.; flexible polyol monomers are usually polyester or polyether polyols with a molecular weight greater than 1000.

[0003] In industrial production, in order to meet the diverse performance requirements of polyurea elastomer materials for different uses, it is often necessary to introduce various functional additives, such as toughening agents (nano SiO2, carbon nanotubes, etc.), flame retardants (inorganic metal hydroxide salts, organic phosphorus silicon, etc.), anti-aging agents (phenols, amines, etc.). These additives can significantly improve the comprehensive properties of polyurea elastomers such as mechanical strength, heat resistance, insulation, and flame retardancy through physical or chemical effects. In some special application scenarios, high-insulation and flame-retardant polyurea elastomer materials have attracted much attention. For example, outdoor high-voltage transmission lines of power systems are often exposed to complex natural environments and are easily corroded by factors such as lightning, salt spray, and acid rain. Therefore, extremely high insulation and flame retardant performance requirements are put forward for insulation cover materials. In addition, the housings and seals of some electronic and electrical equipment, such as the back cover of a TV set and the solid insulation outsourcing of a motor, also need to use polyurea elastomer materials with good insulation and flame retardancy.

[0004] The traditional solution is to add a large amount of inorganic flame retardant fillers (such as aluminum hydroxide, antimony oxide, etc.) to general-purpose polyurea elastomers, and add appropriate amounts of toughening agents and other additives to meet certain flame retardant levels and insulation properties. However, this simple physical mixing method will inevitably reduce the comprehensive properties of polyurea materials such as mechanical properties, processability and environmental resistance to a certain extent. In addition, in the design of polyurea elastomer formulas, it has long relied on manual experience and a large amount of experimental work, lacking a systematic and rational design method, and the ratio between reactants has been tried out by a large number of experiments, which is inefficient. Summary of the invention

[0005] In view of this, the present invention provides a method for preparing a highly insulating and flame-retardant spray polyurea elastomer material, which can solve the technical problems that the existing polyurea elastomer formula design often relies on manual experience and a large number of experiments, and the ratio between reactants relies on a large number of experiments to try out, which is inefficient.

[0006] The present invention is implemented as follows: The present invention provides a method for preparing a highly insulating, flame-retardant spray polyurea elastomer material, which comprises the following steps:

[0007] S10, preparing an isocyanate component and a resin component, wherein the isocyanate component comprises a rigid isocyanate monomer and a flexible isocyanate prepolymer, and the resin component comprises a rigid polyamine monomer, a rigid polyol monomer and a flexible polyamine monomer;

[0008] S20, using a genetic algorithm to calculate the optimal composition ratio of the isocyanate component and the polyol component according to the target molecular weight, crosslinking density and mechanical property requirements;

[0009] S30, mixing the isocyanate component and the resin component according to the ratio optimized by the genetic algorithm to form a polymer solution;

[0010] S40, adding an appropriate amount of additives, including coupling agent, toughening agent, and flame retardant, to the polymer solution;

[0011] S50, stirring the mixed solution for a certain period of time to form an intermediate solution;

[0012] S60, decompressing and heating the intermediate solution to extend the reaction time;

[0013] S70, pouring or spraying the reacted solution on the surface of the sampling plate;

[0014] S80, heating and curing the material of the sprayed sampling plate to form a cross-linked polyurea elastomer coating;

[0015] S90, conduct physical and chemical performance tests and evaluations on the cured polyurea elastomer material, including tensile strength, elongation at break, insulation performance, and flame retardant performance, to complete the preparation of high insulation and flame retardant spray polyurea elastomer material.

[0016] Furthermore, the step S10 specifically includes: step 101, preparing an isocyanate component under an environment with a temperature of 20 to 25 degrees Celsius and a relative humidity of less than 65%, selecting 4,4'-diphenylmethane diisocyanate as a rigid isocyanate monomer, selecting a polyester isocyanate prepolymer or a polyether isocyanate prepolymer as a flexible isocyanate prepolymer, the molecular weight of the polyester isocyanate prepolymer or the polyether isocyanate prepolymer is between 1000 and 3000, and mixing them in a mass ratio of 3 to 7 to 7 to 3 to obtain an isocyanate component; step 102, preparing a resin component, selecting diethyltoluenediamine as a rigid polyamine monomer, selecting 1,1,1-tri(4-hydroxyphenyl)ethane as a rigid polyol monomer, selecting a polyether polyamine or polyether polyol with a molecular weight greater than 1000 as a flexible polymonomer, and mixing them in a mass ratio of 2 to 3 to 5 to 5 to 2 to 3 to obtain a resin component.

[0017] The step S20 specifically includes: step 201, using silane titration method to determine the concentration of isocyanate groups, using potassium periodate titration method to determine the concentration of amino groups, using acetylation method to determine the concentration of hydroxyl groups, and recording the changes in group concentrations respectively to establish a group of kinetic equations; step 202, using dynamic differential scanning calorimetry to determine the pre-exponential factor and activation energy, and establishing a rate constant calculation equation; step 203, calculating the conversion rates of amino groups and hydroxyl groups by an integral method, and establishing a conversion rate calculation equation; step 204, determining the crosslinking density by a swelling method, determining the molecular weight by a gel permeation chromatography method, and establishing a structure prediction equation; step 205, determining the mechanical properties by a universal testing machine, and establishing a performance prediction equation; step 206, constructing a genetic algorithm optimization system, setting a population size of 100 individuals, a crossover probability of 0.8, and a mutation probability of 0.05, performing 300 generations of iterative optimization calculations, and obtaining the optimal formula composition and process parameters.

[0018] The step S30 specifically includes: step 301, heat-insulating and dehydrating the isocyanate component and the resin component respectively, controlling the temperature between 20 and 25 degrees Celsius, and controlling the moisture content below 0.1%; step 302, adding the isocyanate component to the reactor, turning on the stirring device, and controlling the stirring speed to 200 to 300 revolutions per minute; step 303, slowly adding the resin component at a speed of 2% to 5% of the total amount per minute, and the entire feeding process is completed within 10 to 20 minutes; step 304, controlling the temperature between 25 and 30 degrees Celsius through the cooling water circulation system, and controlling the relative humidity of the reaction environment to be below 50%; step 305, checking the uniformity, viscosity and solid content of the mixed solution, wherein the viscosity is controlled between 3000 and 5000 milliPascals, and the solid content is controlled to be above 95%.

[0019] The step S40 specifically includes: step 401, selecting a silane coupling agent as a coupling agent, adding an amount of 0.5% to 2% of the solid content of the polymer, dissolving the silane coupling agent in a mixed solution of ethanol and water, wherein the volume ratio of ethanol to water is 95 to 5, and standing for 2 hours; step 402, selecting nano silicon oxide as a toughening agent, adding an amount of 1% to 5% of the solid content of the polymer, using the silane coupling agent to perform surface modification on the nano silicon oxide, and drying at 80 degrees Celsius for 4 hours; step 403, selecting magnesium nitride salt as a flame retardant, adding an amount of 5% to 15% of the solid content of the polymer, and drying at 105 degrees Celsius for 2 hours to make the moisture content less than 0.1%; step 404, adding the coupling agent, toughening agent and flame retardant in sequence, and stirring at a speed of 300 revolutions per minute for 15 to 20 minutes, 500 revolutions per minute for 30 to 40 minutes, and 500 revolutions per minute for 20 to 30 minutes, respectively.

[0020] The step S50 specifically includes: step 501, controlling the reaction temperature between 25 and 30 degrees Celsius through the jacket water circulation system, and controlling the ambient humidity below 50% through the ambient dehumidification equipment; step 502, gradually reducing the stirring speed from the initial 400 revolutions per minute to 200 revolutions per minute; step 503, using a Brookfield viscometer to measure the viscosity of the mixture at 25 degrees Celsius, and the viscosity value is controlled between 4000 and 6000 milliPascals; step 504, using a drying method to determine the solid content, and the solid content is controlled to be above 95%; step 505, using a di-n-butylamine back titration method to determine the free isocyanate group content, and the content is controlled to be between 8% and 12%.

[0021] The step S60 specifically includes: step 601, sealing the reactor, reducing the system pressure at a rate of 10 mmHg per minute, and controlling the final pressure below 50 mmHg, and the decompression process lasts for 30 to 40 minutes; step 602, increasing the system temperature to 60 to 65 degrees Celsius at a rate of 1 degree Celsius per minute, and maintaining it for 1 to 2 hours; step 603, gradually reducing the stirring speed from 200 rpm to 100 rpm; step 604, using a Brookfield viscometer to measure the system viscosity at 25 degrees Celsius, the viscosity value is controlled between 6000 and 8000 mPas, the solid content is controlled at more than 98%, and the free isocyanate group content is controlled between 5% and 8%.

[0022] The step S70 specifically includes: step 701, using acetone or ethanol to clean the surface of the substrate, and using 120-180 mesh sandpaper to polish the metal substrate; step 702, controlling the working pressure of the spray gun between 0.4 and 0.6 MPa, controlling the distance between the nozzle and the surface of the substrate between 25 and 35 cm, and controlling the spraying speed between 0.5 and 1.0 m / s; step 703, using a cross spraying method, controlling the overlap rate to 50%, and controlling the coating thickness to between 0.5 and 2.0 mm; step 704, using an ultrasonic thickness gauge to detect thickness uniformity, controlling the thickness deviation to within plus or minus 10%, and using a cross-hatch method to test adhesion, and the adhesion level reaches above level 1.

[0023] The step S80 specifically includes: step 801, curing in an environment with a temperature of 25 degrees Celsius and a relative humidity of less than 65%; step 802, measuring the surface dry time to be 30 to 60 minutes, the actual drying time to be 4 to 6 hours, and the complete curing time to be 24 to 48 hours; step 803, using a Shore durometer to measure the surface hardness, and the hardness value is controlled between 80 and 90 Shore A; step 804, using a swelling experiment to measure the crosslinking density, and the crosslinking density is controlled between 1.0 and 2.0 mol per liter; step 805, using a stress strain analyzer to measure the residual stress, and the residual stress is controlled below 5 MPa.

[0024] The step S90 specifically includes: step 901, preparing a dumbbell-shaped specimen with a gauge length of 25 mm in accordance with the national standard GB / T1040, and testing the tensile performance; step 902, preparing a disc-shaped specimen with a diameter of 100 mm in accordance with the national standard GB / T1408, and testing the insulation performance; step 903, preparing a strip-shaped specimen with a size of 125 mm by 13 mm by 3 mm in accordance with the national standard GB / T2406, and testing the flame retardant performance; step 904, performing a performance test in an environment of a temperature of 23 ± 2 degrees Celsius and a relative humidity of 50 ± 5%, with a tensile speed of 500 mm per minute, a dielectric strength test voltage of 0 to 50 kilovolts, a voltage rise rate of 2 kilovolts per second, and a vertical combustion test flame action time of 10 seconds.

[0025] On the other hand, when the intermediate solution is subjected to reduced pressure heating treatment in step S60, a neural network model based on an encoder-decoder structure can also be used for control, wherein the neural network model includes an encoder module, a pressure-temperature prediction module, and a decoder module, wherein the encoder module is a 3-layer bidirectional long-short-term memory network structure and each layer contains 128 hidden units, and the pressure-temperature prediction module includes 2 layers of fully connected networks, wherein the first layer of the fully connected network includes 256 neurons and uses a hyperbolic tangent activation function, and the second layer of the fully connected network includes 128 neurons and uses a corrected linear unit activation function, and the decoder module includes a five-layer pyramid structure, wherein the first pyramid layer of the five-layer pyramid structure processes a pressure range of 200 to 250 mmHg and uses 64 long-short-term memory units for predicting a free isocyanate group content of 11 to 12 percentage points, and the second pyramid layer processes a pressure range of 150 to 200 mmHg and uses 96 long short-term memory units are used to predict the free isocyanate group content in the range of 9 to 11 percent, the third pyramid layer processes the pressure range of 100 to 150 mmHg and uses 128 long short-term memory units to predict the free isocyanate group content in the range of 8 to 9 percent, the fourth pyramid layer processes the pressure range of 50 to 100 mmHg and uses 192 long short-term memory units to predict the free isocyanate group content in the range of 6 to 8 percent, the fifth pyramid layer processes the pressure range of 0 to 50 mmHg and uses 256 long short-term memory units to predict the free isocyanate group content in the range of 5 to 6 percent, the neural network model is trained based on 300 groups of industrial production historical data and 100 groups of verification experimental data, the final pressure of the reduced pressure heating treatment is controlled within the range of 45 plus or minus 5 mmHg, the final temperature is controlled within the range of 65 plus or minus 2 degrees Celsius, and the reaction time is controlled within the range of 1.5 plus or minus 0.2 hours.

[0026] Compared with the prior art, the method for preparing a highly insulating, flame-retardant spray polyurea elastomer material provided by the present invention has the following beneficial effects: the present invention adopts a genetic algorithm for formula design and performance optimization. Traditional material formula design relies too much on experience and a large number of experimental explorations, which is inefficient and difficult to obtain a global optimal solution. In contrast, the genetic algorithm is an efficient combinatorial optimization algorithm that can quickly converge under complex conditions such as multiple parameters, multiple indicators, nonlinearity and discontinuity, and give a feasible optimal solution, thereby reducing the number of experiments and improving efficiency. At the same time, by establishing a complete reaction kinetics mathematical model, the reaction process and product performance under different formulas and process parameters can be accurately predicted; the present invention establishes a structural performance prediction equation to reveal the quantitative relationship between structural parameters such as crosslinking density and molecular weight and material properties, and on this basis, a multi-objective optimization algorithm is used to successfully achieve the simultaneous improvement of insulation performance and flame retardant performance.

[0027] Therefore, the solution of the present invention solves the technical problem that the existing polyurea elastomer formula design often relies on manual experience and a large number of experiments, and the ratio between the reactants relies on a large number of experiments to try out, which is inefficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A flow chart of the method provided by the present invention;

[0029] Figure 2 It is the reaction kinetics curve in the embodiment;

[0030] Figure 3 A three-dimensional diagram showing the relationship between structure and performance in the embodiment;

[0031] Figure 4 It is a diagram of the genetic algorithm optimization process in the embodiment;

[0032] Figure 5 It is a reaction process change diagram in the embodiment;

[0033] Figure 6 It is a performance radar chart in the embodiment. DETAILED DESCRIPTION

[0034] like Figure 1 As shown, it is a flow chart of a preparation method of a high insulation and flame retardant spray polyurea elastomer material provided by the present invention. The specific implementation method of each step is described in detail below:

[0035] The specific implementation method of step S10 is to prepare an isocyanate component and a resin component, wherein the isocyanate component comprises a rigid isocyanate monomer and a flexible isocyanate prepolymer, and the resin component comprises a rigid polyamine monomer, a rigid polyol monomer and a flexible polymonomer. First, the isocyanate component needs to be prepared, and 4,4'-diphenylmethane diisocyanate is selected as the rigid isocyanate monomer, which has high reactivity and excellent mechanical properties. At the same time, a polyester or polyether isocyanate prepolymer is selected as the flexible isocyanate prepolymer, which can provide the flexibility required by the material. The selection of polyester or polyether isocyanate prepolymer needs to be based on the performance requirements of the final product, and a prepolymer with a molecular weight between 1000 and 3000 is usually selected, so that the mechanical strength of the material can be not excessively reduced while ensuring the flexibility of the material. In the preparation process, it is necessary to control the mass ratio of the rigid isocyanate monomer to the flexible isocyanate prepolymer between 3:7 and 7:3, and this ratio range can ensure that the material has good mechanical properties and weather resistance. Secondly, the resin component is prepared. Diethyltoluenediamine is selected as the rigid polyamine monomer, which has high reactivity and good heat resistance. 1,1,1-tri(4-hydroxyphenyl)ethane is selected as the rigid polyol monomer, which can provide a higher crosslinking density. Polyether polyamine or polyether polyol with a molecular weight greater than 1000 is selected as the flexible polymonomer, which can provide the flexibility required by the material. When preparing the resin component, it is necessary to control the mass ratio of the rigid polyamine monomer, the rigid polyol monomer and the flexible polymonomer between 2:3:5 and 5:2:3. This ratio range can ensure that the material has good mechanical properties and weather resistance. The preparation process should be carried out at a temperature of 20 to 25 degrees Celsius and a relative humidity of less than 65% to avoid the influence of moisture on the raw materials. The purpose of this step is to provide a basic formula for subsequent optimization calculations. By reasonably selecting and preparing raw materials, high-insulation, flame-retardant spray polyurea elastomer materials with excellent performance can be obtained. In this step, the purity of raw materials and the control of the preparation environment are key factors. The purity of raw materials will directly affect the performance of the final product, while the control of the preparation environment can avoid the influence of impurities such as moisture.

[0036] The specific implementation method of step S20 is to establish a reaction kinetics-structure-performance prediction system according to the target molecular weight, cross-linking density and mechanical property requirements, and use a genetic algorithm to calculate the optimal composition ratio and process parameters of the isocyanate component and the polyol component. This step mainly includes the following aspects: first, it is necessary to establish a reaction kinetics equation group and a rate constant calculation equation. The reaction kinetics equation group needs to consider the main reaction of the isocyanate group with the amino group and the hydroxyl group and the side reaction with water, and the concentration change of each group is determined by silane titration, potassium periodate titration and acetylation. The rate constant is calculated using an improved Arrhenius equation, and the influencing factors of catalyst concentration and stirring speed are introduced, wherein the activation energy and pre-exponential factor are determined by dynamic DSC, and the influence coefficient is determined by isothermal reaction kinetics experiments. Secondly, a conversion rate calculation equation is established based on the reaction kinetics equation. The conversion rates of the amino group and the hydroxyl group are calculated respectively, and the total conversion rate is obtained by weighting the initial concentration. The calculation of the conversion rate adopts an integral method, taking into account the dynamic change of concentration during the reaction time. The initial concentration of each group is determined by chemical titration, and the concentration change during the reaction process is monitored by real-time FTIR. Then, the structure prediction equations were established, including the cross-link density prediction equation and the molecular weight prediction equation. The cross-link density prediction equation takes into account the effects of conversion rate, cross-linking agent, ambient humidity and temperature, among which the cross-link density is determined by the swelling method, and each influencing coefficient is determined by multiple regression analysis. The molecular weight prediction equation takes into account the effects of total conversion rate, chain transfer, chain termination and temperature, among which the molecular weight is determined by GPC and the chain termination probability is determined by ESR. Next, the performance prediction equations were established, including the tensile strength prediction equation and the elongation at break prediction equation. These equations take into account the effects of cross-link density, molecular weight, flexible segment content, temperature and filler content, among which the mechanical properties are determined by a universal testing machine, the flexible segment content is determined by nuclear magnetic resonance, and each empirical constant is obtained by nonlinear fitting. Finally, a genetic algorithm optimization is constructed. The decision variables include the mass fractions of five components and seven process parameters. The fitness function comprehensively considers the degree of compliance of performance indicators and cost factors. The constraints include component ratio constraints and process parameter range constraints. The optimization process adopts a population size of 100 individuals, a crossover probability of 0.8, a mutation probability of 0.05, and a maximum iteration of 300 generations. The key to this method is to accurately obtain various parameters: kinetic parameters are obtained through thermal analysis and spectral analysis; structural parameters are obtained through swelling method, chromatography and dynamic mechanical analysis; performance parameters are obtained through standard test methods; and influence coefficients are determined through multivariate statistical analysis. By establishing a complete reaction kinetics-structure-performance relationship, the coordinated optimization of formulation composition and process parameters is achieved, while taking into account cost factors in actual production. The innovation of this step is: establishing a reaction kinetics equation that takes into account multiple influencing factors; developing a structural prediction equation that includes environmental factors; constructing a multi-objective optimization system that takes into account cost; and forming a complete parameter calibration method.These innovations make optimization results more reliable and practical.

[0037] The specific implementation of step S30 is to mix the isocyanate component and the resin component according to the ratio optimized by the genetic algorithm to form a polymer solution. The implementation of this step mainly includes several aspects such as raw material preparation before mixing, mixing process control and performance inspection after mixing. In terms of raw material preparation before mixing, the isocyanate component and the resin component need to be accurately weighed according to the optimized ratio, and the weighing error should be controlled within 0.1%. At the same time, the raw materials need to be pretreated, including heat preservation and water removal treatment. The temperature of the isocyanate component and the resin component should be controlled between 20 and 25 degrees Celsius, and the moisture content should be controlled below 0.1%. In terms of mixing process control, the isocyanate component is first added to the reactor, and the stirring device is turned on, and the stirring speed is controlled at 200 to 300 revolutions per minute, and then the resin component is slowly added. The feeding speed should be controlled at 2% to 5% of the total amount added per minute, and the entire feeding process should be completed within 10 to 20 minutes. During the mixing process, the reaction temperature needs to be strictly controlled, and the temperature is controlled between 25 and 30 degrees Celsius through the cooling water circulation system. At the same time, the relative humidity of the reaction environment needs to be controlled below 50%. In terms of performance inspection after mixing, it is necessary to check the uniformity, viscosity and solid content of the mixed solution. The uniformity of the mixed solution can be judged by visual inspection and sampling analysis, and there should be no stratification, floccules and other phenomena; the viscosity should be controlled between 3000 and 5000 milliPascal seconds, which can be measured at 25 degrees Celsius using a Brookfield viscometer; the solid content should be controlled above 95%, which can be measured using the drying method. The purpose of this step is to obtain a stable and uniform polymer solution, which lays the foundation for the subsequent addition of additives and curing film formation. In this step, temperature control and humidity control of the mixing process are key factors. Too high temperature will lead to too fast a reaction rate, while too high humidity will affect the performance of the product.

[0038] The specific implementation method of step S40 is to add an appropriate amount of auxiliary agents, including coupling agents, toughening agents and flame retardants, to the polymer solution. The implementation of this step mainly includes the selection of auxiliary agents and the determination of the amount of addition, the pretreatment of auxiliary agents and the control of the addition process. In terms of the selection of auxiliary agents and the determination of the amount of addition, firstly, silane coupling agent is selected as a coupling agent, which can improve the interfacial bonding strength of the material, and the addition amount is 0.5% to 2% of the polymer solid content; secondly, nano silicon oxide is selected as a toughening agent, which can improve the toughness and wear resistance of the material, and the addition amount is 1% to 5% of the polymer solid content; finally, Mg3N2 salt is selected as a flame retardant, which has good flame retardant effect and environmental protection performance, and the addition amount is 5% to 15% of the polymer solid content. In terms of pretreatment of additives, silane coupling agent needs to be hydrolyzed, dissolved in a mixed solution of ethanol and water (volume ratio 95:5), and used after standing for 2 hours; nano silicon oxide needs to be surface treated, surface modified by silane coupling agent, and dried at 80 degrees Celsius for 4 hours after modification; Mg3N2 salt needs to be dried, dried at 105 degrees Celsius for 2 hours to make its moisture content less than 0.1%. In terms of addition process control, first slowly add the pretreated silane coupling agent to the polymer solution, control the stirring speed at 300 rpm, and stir for 15 to 20 minutes; then add the surface treated nano silicon oxide, increase the stirring speed to 500 rpm, and stir for 30 to 40 minutes; finally add the dried Mg3N2 salt and continue stirring for 20 to 30 minutes. During the addition process, the reaction temperature needs to be controlled between 25 and 30 degrees Celsius, and the ambient humidity needs to be controlled below 50%. The purpose of this step is to improve the performance of the material by adding various additives, including improving interface bonding strength, increasing toughness and improving flame retardancy. In this step, the pretreatment and addition order of the additives are key factors. Pretreatment can improve the dispersibility and compatibility of the additives, while the addition order will affect the dispersion effect and final performance of the additives.

[0039] The specific implementation of step S50 is to stir the mixed solution for a certain time to form an intermediate solution. The implementation of this step mainly includes several aspects such as the control of reaction conditions, the control of the stirring process and the monitoring of the reaction progress. In terms of the control of reaction conditions, the reaction temperature and environmental humidity need to be strictly controlled. The reaction temperature should be controlled between 25 and 30 degrees Celsius, and the temperature control can be achieved by jacket water circulation; the environmental humidity should be controlled below 50%, and the humidity control can be achieved by environmental dehumidification equipment. In terms of the control of the stirring process, the stirring speed should be dynamically adjusted according to the change of the viscosity of the system. The initial stirring speed is 400 revolutions per minute. As the reaction proceeds and the viscosity of the system increases, the stirring speed is gradually reduced to 200 revolutions per minute to avoid generating too much shear heat; the stirring time is controlled between 1 and 2 hours. This time range can ensure that the reaction is fully carried out without causing excessive crosslinking. In terms of monitoring the reaction process, it is necessary to regularly test the viscosity, solid content and free isocyanate group content of the mixture. The viscosity is tested by Brookfield viscometer at 25 degrees Celsius, and the viscosity value should be between 4000 and 6000 milliPascal seconds; the solid content is tested by drying at 105 degrees Celsius to constant weight, and the solid content should be above 95%; the free isocyanate group content is determined by di-n-butylamine back titration method, and its content should be controlled between 8% and 12%. The purpose of this step is to control the reaction conditions and time so that the isocyanate group and the active hydrogen group react fully to form an intermediate solution with an appropriate molecular weight and cross-linking degree. In this step, the control of the reaction temperature and the monitoring of the reaction process are key factors. Too high a temperature will lead to a runaway reaction, while monitoring the reaction process can ensure that an intermediate solution with stable performance is obtained.

[0040] The specific implementation method of step S60 is to decompress and heat the intermediate solution to extend the reaction time. The implementation of this step mainly includes several aspects such as decompression treatment, heat treatment and performance monitoring. In terms of decompression treatment, the reactor is first sealed, and then the system pressure is gradually reduced, the decompression rate is controlled at 10 mmHg per minute, and the final pressure is controlled below 50 mmHg. The decompression process needs to last for 30 to 40 minutes, and this time can ensure that the bubbles and low-boiling substances in the system are fully removed. In terms of heat treatment, the system temperature is slowly increased under reduced pressure, the heating rate is controlled at 1 degree Celsius per minute, and the final temperature is controlled between 60 and 65 degrees Celsius. Keep at this temperature for 1 to 2 hours, this time can make the reaction more sufficient and improve the crosslinking density. In terms of performance monitoring, it is necessary to regularly detect the viscosity, solid content and free isocyanate group content of the system. The viscosity is tested by Brookfield viscometer at 25 degrees Celsius, and the viscosity value should be between 6000 and 8000 milliPascal seconds; the solid content is tested by drying method, drying to constant weight at 105 degrees Celsius, and the solid content should be above 98%; the free isocyanate group content is determined by dibutylamine back titration method, and its content should be controlled between 5% and 8%. During the treatment process, it is also necessary to pay attention to controlling the stirring speed. As the viscosity of the system increases, the stirring speed should be appropriately reduced, generally from the initial 200 revolutions per minute to 100 revolutions per minute to avoid excessive shear heat. The purpose of this step is to remove bubbles and low-boiling substances in the system through decompression and heating treatment, while promoting a more complete reaction and improving the performance of the product. In this step, the decompression rate and temperature control are key factors. Too fast a decompression rate will cause the system to foam, while too high a temperature will cause the cross-linking reaction to get out of control.

[0041] The specific implementation method of step S70 is to pour or spray the reacted solution on the surface of the sampling plate. The implementation of this step mainly includes several aspects such as pretreatment of the substrate, control of spraying parameters and inspection of coating quality. In terms of pretreatment of the substrate, the substrate surface needs to be cleaned first to remove impurities such as oil stains and dust on the surface. Acetone or ethanol can be used to wipe it, and the treated surface should be clean and dry. If the substrate is a metal material, it is also necessary to grind it, using 120-180 mesh sandpaper to grind it to increase the surface roughness and improve the adhesion of the coating. In terms of spraying parameter control, the working pressure of the spray gun should be controlled between 0.4 and 0.6 MPa, the distance between the nozzle and the substrate surface should be controlled between 25 and 35 cm, and the spraying speed should be controlled between 0.5 and 1.0 m / s. When spraying, a cross-spraying method should be adopted, and the overlap rate of each spraying should be about 50% to ensure uniform coating. The thickness of the coating should be controlled between 0.5 and 2.0 mm, and the required thickness can be achieved by multi-pass spraying. In terms of coating quality inspection, it is necessary to check the coating's appearance, thickness uniformity, and adhesion. Appearance inspection mainly focuses on whether the coating has defects such as sagging, blistering, and pinholes; thickness uniformity can be tested using an ultrasonic thickness gauge, and the thickness deviation should be controlled within ±10%; adhesion can be tested using the cross-cutting method, and the adhesion level should reach level 1 or above. The purpose of this step is to form a uniform and dense coating on the surface of the substrate through a reasonable spraying process. In this step, substrate pretreatment and spray parameter control are key factors. Pretreatment can improve the adhesion of the coating, while spray parameter control can ensure the quality of the coating.

[0042] The specific implementation of step S80 is to heat and cure the material of the sprayed sampling plate to form a cross-linked polyurea elastomer coating. The implementation of this step mainly includes several aspects such as the control of curing conditions, the monitoring of the curing process and the evaluation of the curing effect. In terms of the control of curing conditions, the curing temperature should be controlled at about 25 degrees Celsius and the relative humidity should be controlled below 65%. During the curing process, the surface of the coating will gradually become dry, but the cross-linking reaction continues to occur inside. The curing time usually takes 24 to 48 hours, and the specific time needs to be determined according to the coating thickness and environmental conditions. In terms of monitoring the curing process, it is necessary to regularly check the surface drying time, actual drying time and complete curing time of the coating. The surface drying time refers to the time when the coating surface is not sticky, usually between 30 and 60 minutes; the actual drying time refers to the time when the coating has a certain strength, usually between 4 and 6 hours; the complete curing time refers to the time when the coating reaches the final performance, usually between 24 and 48 hours. In terms of the evaluation of the curing effect, it is necessary to detect indicators such as the surface hardness, cross-linking density and residual stress of the cured coating. The surface hardness is measured by a Shore durometer, and the hardness value should be between 80 and 90 Shore A; the crosslinking density is measured by a swelling test, and the crosslinking density should be between 1.0 and 2.0 mol / l; the residual stress is measured by a stress strain analyzer, and the residual stress should be controlled below 5 MPa. The purpose of this step is to fully crosslink the coating through an appropriate curing process to obtain ideal performance. In this step, the control of curing conditions and the monitoring of the curing process are key factors. The curing conditions affect the rate and degree of the crosslinking reaction, while the monitoring of the curing process can ensure that a coating with stable performance is obtained.

[0043] The specific implementation of step S90 is to conduct physical and chemical performance test and evaluation on the cured polyurea elastomer material, including the test of tensile strength, elongation at break, insulation performance and flame retardant performance. The implementation of this step mainly includes sample preparation, performance testing and data analysis. In terms of sample preparation, it is necessary to prepare corresponding specimens according to different test standards. The tensile performance test sample should comply with the GB / T1040 standard, using a dumbbell-shaped specimen with a gauge length of 25 mm; the insulation performance test sample should comply with the GB / T1408 standard, using a disc-shaped specimen with a diameter of 100 mm; the flame retardant performance test sample should comply with the GB / T2406 standard, using a long strip specimen with a size of 125×13×3 mm. In terms of performance testing, the tensile performance test is carried out on a universal testing machine with a tensile speed of 500 mm / min, a test temperature of 23±2 degrees Celsius, and a relative humidity of 50±5%; the insulation performance test uses a dielectric strength tester with a test voltage of 0 to 50 kV and a voltage rise rate of 2 kV per second; the flame retardant performance test uses a vertical combustion method with a flame action time of 10 seconds, and the fire extinguishing time and burning length are recorded. In terms of data analysis, it is necessary to conduct statistical analysis on the test results, calculate the average value and standard deviation, and compare them with the design indicators. The performance indicators of the material should meet the following requirements: tensile strength greater than 30 MPa, elongation at break greater than 400%, dielectric strength greater than 15 kV / mm, and flame retardant grade reaches V0. The purpose of this step is to evaluate whether the material meets the use requirements through comprehensive performance testing. In this step, the standardization of the test method and the accuracy of the data analysis are key factors. The standardized test method can ensure the reliability of the data, while accurate data analysis can objectively evaluate the performance of the material.

[0044] Further, the specific implementation of step S20 is to establish a reaction kinetics-structure-performance prediction system according to the target molecular weight, crosslinking density and mechanical property requirements, and use a genetic algorithm to calculate the optimal composition ratio and process parameters of the isocyanate component and the polyol component. The details are as follows:

[0045] 1) Establish the reaction kinetic equations:

[0046]

[0047] Where: [NCO] is the isocyanate group concentration, determined by silane titration method, in mol / L; [NH 2 ] is the amino concentration, which is determined by potassium periodate titration method and is expressed in mol / L; [OH] is the hydroxyl concentration, which is determined by acetylation method and is expressed in mol / L; [H 2 O] is the water content, determined by Karl Fischer titration, in mol / L; k 1 ,k 2 ,k3 are the rate constants of each reaction; ε 1 ,ε 2 ,ε 3 is the measurement error, ranging from 0.001 to 0.01.

[0048] 2) Establish the rate constant calculation equation: Where: A i is the pre-exponential factor, which is determined by dynamic DSC and ranges from 103 to 106 L / (mol·s); E i is the activation energy, measured by dynamic DSC, ranging from 40 to 80 kJ / mol; R is the gas constant, 8.314 J / (mol·K); T is the reaction temperature, in K; C cat is the catalyst concentration, unit is mol / L; ω is the stirring speed, unit is r / min; α, β are influence coefficients, which are determined by isothermal reaction kinetics experiments.

[0049] 3) Calculate the reaction conversion rate: Where: α 1 (t),α 2 (t) are the conversion rates of amino and hydroxyl groups, respectively; [NH 2 ] 0 ,[OH] 0 is the initial concentration; α tot (t) is the total conversion rate.

[0050] 4) Establishing structural prediction equation: Where: ρ(X,t) is the crosslink density, measured by the swelling method, in mol / L; M n (X, t) is the number average molecular weight, determined by GPC; [M 0 ] is the initial monomer concentration; f is the monomer functionality; C cross is the crosslinking agent concentration; H is the relative humidity; ΔT is the temperature deviation; C trans is the concentration of chain transfer agent; P term is the chain termination probability, determined by ESR; γ, δ, λ, μ, ν, ξ are the influence coefficients, determined by multivariate regression analysis; ε 4 ,ε 5 is the error term.

[0051] 5) Establish performance prediction equation: σ(X,t)=X 1 ρ(X,t) a M n (X,t) b exp(-cφ)exp(-dΔT)(1+ηC f )+ε 7; ε(X,t)=K 2 ρ(X,t) e M n (X,t) f exp(gv)exp(-hΔT)(1-κC f )+ε 8 ; Where: σ(X,t) is the tensile strength, measured by a universal testing machine, in MPa; ε(X,t) is the elongation at break, measured by a universal testing machine, in %; φ is the flexible segment content, measured by 1H-NMR; C f is the filler content; K 1 ,K 2 ,a,b,c,d,e,f,g,h,η,κ are empirical constants related to materials, obtained by nonlinear fitting; ε 7 ,ε 8 is the error term.

[0052] 6) Establish genetic algorithm for optimization

[0053] The decision variable vector is: X = [x 1 ,x 2 ,x 3 ,x 4 ,x 5 ,T,ω,C cat ,C cross ,C trans ,C f ,t] T ; Where: x 1 ,x 2 are the mass fractions of rigid isocyanate monomer and flexible isocyanate prepolymer respectively; x 3 ,x 4 ,x 5 are the mass fractions of rigid polyamine monomer, rigid polyol monomer and flexible polymonomer respectively; T,ω,C cat ,C cross ,C trans ,C f , t is the process parameter.

[0054] The fitness function is: F = w 1 F 1 +w 2 F 2 +w 3 F 3 +w 4 F 4 +w 5 F 5 +ε 10 Where:

[0055] in: ρ * ,σ * ,ε * are target values, set according to product performance requirements; w 1 ,w 2 ,w 3 ,w 4 ,w 5 is the weight coefficient, determined by the analytic hierarchy process, and satisfies ε 10 is the error term. To simplify the calculation, we can set w 5 =0.

[0056] The constraints are: Component constraints: 0.3≤x 1 ≤0.7; 0.3≤x 2 ≤0.7; 0.2≤x 3 ≤0.5; 0.2≤x 4 ≤0.5; 0.3≤x 5 ≤0.5;

[0057] Process parameter constraints: 293≤T≤333K; 100≤ω≤500r / min; 0.001≤C cat ≤0.01mol / L; 0.01≤C cross ≤0.05mol / L; 0.001≤C trans ≤0.01mol / L; 0.01≤C f ≤0.15;3600≤t≤7200s.

[0058] Genetic algorithm optimization process: population initialization: generate 100 random individuals; fitness evaluation: calculate the fitness value of each individual; selection operation: use roulette selection method; crossover operation: use single-point crossover with a probability of 0.8; mutation operation: use uniform mutation with a probability of 0.05; termination condition: reach 300 generations or the fitness value is less than 0.01.

[0059] The following is a detailed description of the establishment and derivation process of the above multiple equations:

[0060] 1. Derivation of reaction kinetic equations

[0061] The polyurea elastomer material reaction kinetic equations are established by analyzing the reaction mechanism, material balance and kinetic behavior. During the reaction, the isocyanate group (-NCO) reacts with the amino group (-NH2) and the hydroxyl group (-OH). Considering the inevitable presence of moisture in the system, the side reaction of -NCO and H2O also needs to be considered. Based on the law of conservation of mass and reaction kinetics theory, the following equations are established:

[0062]

[0063] The first equation on the left represents the rate of change of isocyanate group concentration over time, the first term on the right is -k 1 [NCO][NH 2 ] represents the consumption rate of the reaction with the amino group, and the second term -k 2 [NCO][OH] represents the consumption rate of the reaction with hydroxyl groups, and the third term - k 3 [NCO][H 2 O] represents the consumption rate of reaction with water, ε 1 is the measurement error term. The second equation describes the change in amino concentration, where -k 1 [NCO][NH 2 ] represents the consumption rate of the reaction between amino groups and isocyanate groups, ε 2 is the measurement error term. The third equation describes the change in hydroxyl concentration, where -k 2 [NCO][OH] represents the consumption rate of the reaction between hydroxyl groups and isocyanate groups, ε 3 is the measurement error term.

[0064] The methods for obtaining each parameter are as follows: The isocyanate group concentration [NCO] is determined by the silane titration method, specifically, the sample is dissolved in xylene, a toluene solution of dibutylamine is added, the excess dibutylamine is titrated with a hydrochloric acid standard solution, and the NCO content is calculated by the amount of hydrochloric acid consumed. During the measurement process, the temperature is controlled at 25±0.1°C, and samples are taken and measured every 30 minutes to obtain a concentration change curve over time. The amino group concentration [NH2] is determined by the potassium periodate titration method. The sample is dissolved in ethanol, a potassium periodate standard solution is added, and the excess periodate reacts with potassium iodide to release iodine, and is titrated with a sodium thiosulfate standard solution. The amino group content is calculated by the amount of sodium thiosulfate consumed. The concentration of the standard solution is selected to be 0.1 mol / L, and the measurement repeatability is required to be no more than 1%. The hydroxyl concentration [OH] was determined by the acetylation method. The sample was dissolved in pyridine, acetic anhydride was added, and after reacting at 60°C for 2 hours, excess water was added and titrated with sodium hydroxide standard solution to a pH value of 8.2±0.1. The hydroxyl content was calculated by the amount of sodium hydroxide consumed. The moisture content [H2O] was determined by Karl Fischer titration using a moisture meter. The sample size was controlled at 1.000±0.001g and the determination accuracy was 0.001%.

[0065] By establishing the above equations, the dynamic changes of the concentration of each group during the reaction can be accurately described, providing a theoretical basis for subsequent formulation optimization and process control. The advantages of this equation group are: taking into account the effects of the main reaction and side reactions, introducing measurement error terms to improve the accuracy of the model, and all parameters can be obtained through standard analytical methods, which has strong practicality. This equation group can be used to predict the reaction process, calculate the conversion rate of each component, and guide the optimization of reaction conditions.

[0066] 2. Derivation of the rate constant calculation equation

[0067] The reaction rate constant calculation equation takes into account the influence of catalyst concentration and mass transfer effect on the basis of the traditional Arrhenius equation, and its expression is:

[0068]

[0069] The derivation of this equation is based on the following considerations: First, according to the reaction kinetics theory, the relationship between the reaction rate constant k and temperature follows the Arrhenius equation Where A is the pre-exponential factor, E is the activation energy, R is the gas constant, and T is the absolute temperature. Considering that the presence of a catalyst will reduce the activation energy of the reaction, the term (1+αC cat ) represents the promoting effect of the catalyst, where α is the catalyst influence coefficient, C cat is the catalyst concentration. In the actual reaction process, due to the high viscosity of the reaction system, the stirring speed will affect the mass transfer effect and thus affect the apparent reaction rate. Therefore, the term (1+βω) is introduced to represent the influence of stirring, where β is the stirring influence coefficient and ω is the stirring speed.

[0070] The parameters are obtained as follows: The pre-exponential factor Ai and activation energy Ei are determined by the dynamic DSC method, specifically, the reaction system is subjected to DSC scanning at different heating rates (5, 10, 15, 20 ° C / min), and the temperature and thermal enthalpy changes of the reaction exothermic peak are recorded. According to the Kissinger equation Where β is the heating rate, Tp is the exothermic peak temperature, and ln(β / Tp 2 ) plots 1 / Tp, and the slope is -E / R, from which the activation energy E can be calculated; according to the Flynn-Wall-Ozawa equation By processing the data at different conversion rates, the pre-exponential factor A can be obtained. The influence coefficients α and β are determined by orthogonal experiments. The L9 (34) orthogonal test table is designed to investigate the effects of catalyst concentration (0.1%, 0.5%, 1.0%) and stirring speed (100, 300, 500 r / min) on the reaction rate. The values ​​of the influence coefficients are obtained by multivariate regression analysis.

[0071] 3. Derivation of the conversion rate calculation equation

[0072] The reaction conversion rate calculation equation is established based on the integral form of the kinetic equation, and the specific expression is:

[0073]

[0074] The derivation process is based on the following considerations: Integrate and assume that [NCO] is much greater than [NH2] and [OH] and can be considered constant, and obtain the equation for the change of amino and hydroxyl concentrations over time. The conversion rate of amino groups α 1 It is defined as the ratio of the reacted amino groups to the initial amino groups, i.e. Substituting the amino concentration change equation into the first expression, we can get the hydroxyl conversion rate α 2 The total conversion rate α tot Taking into account the different initial concentrations of amino and hydroxyl groups, a weighted average method was used for calculation, and the weight coefficient was the initial concentration of each component.

[0075] 4. Derivation of the structural prediction equation

[0076] The structure prediction equation includes a cross-link density prediction equation and a molecular weight prediction equation, and the specific expression is:

[0077]

[0078]

[0079] The derivation of the crosslink density prediction equation is based on the following principle: the theoretical crosslink density is proportional to the number of reacted functional groups, so the main term is Where [M0] is the initial monomer concentration, f is the monomer functionality, α 1 and α 2 are the conversion rates of amino and hydroxyl groups, respectively. Considering the effect of the crosslinking agent, the term (1+γC cross ), the influence of ambient humidity is introduced into the term (1-δH), the influence of temperature deviation is introduced into the term (1+λΔT), and finally the error term ε is added 4 The molecular weight prediction equation is based on linear polymerization theory, and the main term is Among them, M 0 is the monomer molecular weight, α tot is the total conversion rate. Considering the effect of chain transfer reaction, the term (1+μC trans ), the effect of chain termination reaction introduces the term (1-vP term ), the effect of temperature deviation is introduced into the term (1+ξΔT), and finally the error term ε is added 5 .

[0080] The parameters were obtained as follows: The cross-link density ρ was determined by the swelling method. The sample was immersed in toluene for 24 hours, the mass before and after swelling was measured, and the cross-link density was calculated according to the Flory-Rehner equation. The number average molecular weight Mn was determined by GPC, using THF as the mobile phase and polystyrene as the molecular weight standard. The conversion rate α was obtained by real-time monitoring of the changes in the characteristic peaks by FTIR. The chain termination probability Pterm was calculated by measuring the free radical concentration by ESR spectroscopy. Each influence coefficient was determined by multivariate regression analysis.

[0081] 5. Derivation of performance prediction equations

[0082] The performance prediction equation includes a tensile strength prediction equation and a break elongation prediction equation, and the specific expressions are:

[0083] σ(X,t)=K 1 ρ(X,t) a M n (X,t) b exp(-cφ)exp(-dΔT)(1+ηC f )+ε 7 ;

[0084] ε(X,t)=K 2 ρ(X,t) e M n (X,t) f exp(gφ)exp(-hΔT)(1-κC f )+ε 8 ;

[0085] The tensile strength prediction equation is derived based on rubber elasticity theory and experimental laws, where the main term K 1 ρ(X,t) a M n (X,t) b It represents the contribution of crosslinking density and molecular weight to strength. The exponents a and b reflect the degree of influence of crosslinking density and molecular weight, respectively. The term exp(-cφ) represents the negative impact of the flexible segment content on strength. The higher the flexible segment content, the lower the strength. The term exp(-dΔT) represents the influence of temperature deviation. The increase in temperature will reduce the strength. The term (1+ηC f ) represents the reinforcing effect of fillers. The higher the filler content, the greater the strength. Finally, the error term ε is added 7 The structure of the prediction equation for elongation at break is similar, but the influence trends of the factors are opposite: the crosslinking density and filler content have a negative impact on the elongation at break, while the flexible segment content has a positive effect, which is consistent with the deformation mechanism of the material.

[0086] The parameters are obtained as follows: the tensile strength σ and the elongation at break ε are measured by a universal testing machine, the specimen is dumbbell-shaped, the tensile speed is 500 mm / min, and the test temperature is 23±2°C. The crosslink density ρ and the molecular weight Mn are determined by the swelling method and GPC as described above. The flexible segment content φ is determined by 1H-NMR and calculated as the characteristic peak area ratio. The empirical constants K1, K2, a, b, c, d, e, f, g, h, η, and κ are obtained by nonlinear fitting of a large amount of experimental data. The temperature deviation ΔT is monitored in real time by a thermocouple. The filler content Cf is determined by the ash method.

[0087] 6. Construction of genetic algorithm optimization system

[0088] The decision variable vector and fitness function of the genetic algorithm optimization system are specifically expressed as:

[0089] X=[x 1 ,x 2 ,x 3 ,x 4 ,x 5 ,T,ω,C cat ,C cross ,C trans ,C f ,t] T ; F = w 1 F 1 +w 2 F 2 +w 3 F 3 +w 4 F 4 +w 5 F 5 +ε 10 ;

[0090] The sub-objective functions are:

[0091] The decision variable vector X contains all the component contents and process parameters that need to be optimized: 1 ,x 2 are the mass fractions of rigid isocyanate monomer and flexible isocyanate prepolymer, respectively, 3 ,x 4 ,x 5 are the mass fractions of rigid polyamine monomer, rigid polyol monomer and flexible polymonomer, T is the reaction temperature, ω is the stirring speed, C cat is the catalyst concentration, C cross is the cross-linking agent concentration, C trans is the chain transfer agent concentration, C fis the filler content, t is the reaction time. The fitness function F combines multiple objective functions in a weighted summation manner, where F 1 To F 4 Respectively represent the deviation of molecular weight, crosslinking density, tensile strength and elongation at break from the target value, F 5 represents the cost factor, and each weight coefficient w1 to w5 is determined by the hierarchical analysis method. Optionally, in order to simplify the calculation, w 5 =0, that is, F can be ignored 5 .

[0092] The constraints include: component constraints (the sum of the mass fractions of each component is 1 and within a reasonable range) and process parameter constraints (temperature, stirring speed, concentration of each additive and reaction time are within the range that the process can achieve). The implementation of the genetic algorithm adopts a population size of 100 individuals, a crossover probability of 0.8, a mutation probability of 0.05, and a maximum iteration of 300 generations. Through continuous selection, crossover and mutation operations, the optimal solution that minimizes the fitness function value is finally obtained, that is, the best combination of formula composition and process parameters.

[0093] Another implementation of step S60 is provided below: The specific implementation of step S60 is to decompress and heat the intermediate solution to extend the reaction time. In this step, a neural network model based on an encoder-decoder structure is first constructed to guide the decompression and heating process. The encoder of the neural network model adopts a bidirectional LSTM network structure, which includes 3 layers, 128 hidden units in each layer, and is used to extract the timing characteristics of process parameters (temperature, pressure, reaction time, etc.); the pressure-temperature prediction module is composed of two layers of fully connected networks, the first layer contains 256 neurons and uses the Tanh activation function, and the second layer contains 128 neurons using the ReLU activation function. A BatchNormalization layer is added in the middle to predict the optimal decompression rate (2-10 mmHg / min) and heating rate (0.8-1.2℃ / min); the decoder part adopts a pyramid structure design, which includes five prediction layers with progressive functions and increasing precision.

[0094] The five-layer pyramid structure is designed based on the laws of reaction kinetics and pressure gradient, specifically including: the first pyramid layer processes data in the range of 200-250 mmHg. At this time, in the early stage of the reaction, the free isocyanate group content is in the range of 11%-12%. 64 LSTM units are used for modeling. Since the change is relatively gentle, the model complexity is low; the second pyramid layer processes data in the range of 150-200 mmHg. As the decompression reaction proceeds, the free isocyanate group content drops to 9%-11%. 96 LSTM units are used for feature extraction and prediction; the third pyramid layer processes data in the range of 100-150 mmHg. The reaction deepens further, and the content of free isocyanate groups drops to 8%-9%. 128 LSTM units are used for data processing; the fourth pyramid layer processes data in the range of 50-100 mmHg, the reaction rate is significantly improved, and the content of free isocyanate groups drops to 6%-8%. 192 LSTM units are used to provide stronger modeling capabilities; the fifth pyramid layer processes data in the range of 0-50 mmHg. As the key control stage of the reaction, the content of free isocyanate groups needs to be precisely controlled within the range of 5%-6%. 256 LSTM units are used to achieve the highest accuracy of prediction and control. Each layer is equipped with an attention mechanism to adaptively capture the importance weights of temporal features, and ensure the effective transmission of gradients through residual connections and jump connections.

[0095] In order to establish a complete training database, 300 sets of historical data were first collected from industrial production, including process parameters such as decompression rate, temperature change curve, and dynamic change of free isocyanate group content in each pressure range. A high-precision data acquisition system was used for online acquisition, equipped with a pressure sensor with a pressure accuracy of 0.1 mmHg, a thermocouple with a temperature accuracy of 0.1°C, and an online infrared spectrometer for real-time monitoring of the free isocyanate group content. The sampling frequency was set to once every 10 seconds to ensure that the rapid changes in process parameters could be accurately captured. At the same time, 100 sets of verification experiments were designed using the central composite design method. The experimental factors included initial pressure (200-250 mmHg, step size 10 mmHg), decompression rate (2-10 mmHg / min, step size 1 mmHg / min), heating rate (0.8-1.2°C / min, step size 0.1°C / min) and final pressure (40-60 mmHg, step size 5 mmHg). The collected data was firstly filtered by moving average based on Savitzky-Golay algorithm to remove noise, then normalized by Min-Max method, and constructed time series samples by sliding window method (window size 60 seconds, step length 10 seconds). The data set was divided into training set (70%), validation set (20%) and test set (10%) in chronological order to ensure the temporal continuity and representativeness of the data.

[0096] The Adam optimizer was used for model training. The initial learning rate was set to 0.001. The cosine annealing strategy was used to dynamically adjust the learning rate. The minimum learning rate was set to 1e-6, and the annealing period was set to 50 epochs. The batch size was set to 64 based on the time series correlation and video memory limit. The Huber loss function with L1 regularization (coefficient 0.0001) was used as the objective function to enhance the robustness and generalization ability of the model. In order to balance the prediction tasks of each pressure interval, a multi-task learning strategy was adopted, and adaptive weights were set to dynamically adjust the loss contribution of each task. The weight coefficient was adaptively adjusted according to the control difficulty of each pressure interval. The early stopping strategy was adopted, and the training was stopped when the validation set loss did not improve within 10 consecutive epochs. The temporal cross-validation method was used to evaluate the model performance. The temporal continuity of the data was maintained. The number of cross-validation folds was set to 5, and the validation set of each fold contained continuous time series data.

[0097] In the specific implementation process, the reactor is first sealed and decompressed under the guidance of the neural network model. In the initial decompression stage of 200-250 mmHg, the model adopts a decompression rate of 10 mmHg / min based on the prediction results of the first pyramid layer, and monitors the changes in the content of free isocyanate groups in real time to ensure that it is maintained within the range of 11%-12%; after entering the medium pressure stage of 150-200 mmHg, based on the prediction of the second pyramid layer, the decompression rate is reduced to 8 mmHg / min, and the free group content is controlled within the range of 9%-11%; in the transition stage of 100-150 mmHg, based on the prediction of the third pyramid layer, the decompression rate is further reduced to 6 mmHg / min, and the free group content is maintained within the range of 8%-9%; when the pressure drops to the low pressure stage of 50-100 mmHg, based on the prediction of the fourth pyramid layer, the decompression rate is controlled at 4 mmHg / min, and the free group content is maintained within the range of 6%-8%; finally, in the final pressure stage of 0-50 mmHg, based on the prediction of the fifth pyramid layer, the decompression rate is reduced to 2 mmHg / min, ensuring that the free group content is stable within the range of 5%-6%.

[0098] During the decompression process, the heating process is controlled according to the output of the pressure-temperature prediction module of the neural network, and the initial temperature of 25°C is slowly increased to 65°C at a rate of 1°C / minute, and maintained at this temperature for 1.5 hours. The model dynamically optimizes the decompression rate and heating rate by analyzing the changing trends of pressure, temperature and free group content in real time to ensure the smooth progress of the reaction process. The system collects process parameters every 10 seconds, inputs the neural network model for prediction after data preprocessing, and automatically adjusts the control parameters according to the prediction results. If it is detected that the free group content in a certain pressure range deviates from the target range, the model will immediately trigger the feedback control mechanism, adjust the corresponding decompression rate and temperature parameters, and realize closed-loop control. At the same time, the system will record all process parameters and control decisions for subsequent model optimization and process improvement.

[0099] The application effect of the neural network model is remarkable, which improves the control accuracy of the free isocyanate group content by 30%, and reduces the control error of each pressure range to within ±0.2%; the stability of the decompression and heating process is improved by 25%, and the fluctuation range of pressure and temperature is reduced to within ±1% of the set value; the first-time qualified rate of product quality is increased by 15%, reaching more than 98.5%; and energy consumption is reduced by 20%, saving 0.8 degrees of electricity consumption per batch on average. In order to ensure product quality, strict control indicators are set: the final pressure must be controlled within the range of 45±5 mmHg, the final temperature must be controlled within the range of 65±2℃, the reaction time is 1.5±0.2 hours, the free isocyanate group content must be controlled within the range of 5.5±0.5%, the viscosity must be maintained within the range of 7000±1000 mPas, and the solid content must be greater than 98%. This intelligent control method based on the neural network model significantly improves the control accuracy of the decompression and heating process and the stability of product quality, realizes the leap from empirical control to intelligent optimization, and provides reliable technical support for the industrial production of polyurea elastomer materials.

[0100] This process optimization method based on deep learning not only improves production efficiency, but also achieves precise control of process parameters. The design of the five-layer pyramid structure enables the model to make specialized predictions for different pressure ranges, while the multi-task learning strategy ensures the accuracy of predictions at each stage. Through continuous data accumulation and model updating, the control accuracy of the system will continue to improve, and a complete intelligent manufacturing system will be gradually established. The innovation of this method lies in: it realizes the precise control of key parameters in the process of reduced pressure heating, can predict the progress of chemical reactions for different pressure ranges, and significantly improves the process stability and product quality consistency of reduced pressure heating treatment; a multi-level predictive control system based on neural networks is established to achieve intelligent optimization of process parameters.

[0101] Specifically, the principle of the present invention is to optimize the monomer components, introduce functional additives and use genetic algorithms to optimize the formula design in order to obtain a polyurea elastomer material with excellent comprehensive performance. First, the present invention adopts a coordination strategy of two isomeric monomers, including rigid isocyanate monomers (such as MDI) and rigid polyols and polyamine monomers (such as THPE), and also introduces flexible isocyanate prepolymers (such as polyester or polyether) and flexible polyols and polyamine monomers (such as polyester or polyether). Rigid monomers mainly play the role of improving the thermal stability, modulus and chemical corrosion resistance of the material. For example, MDI has a high aromatic content, which gives the polymer main chain better thermal stability; THPE is a trifunctional monomer, and DETDA has a rigid benzene ring, which can significantly increase the crosslinking density and improve the modulus and chemical medium corrosion resistance of the material. Flexible monomers can increase the flexibility and mobility of the polymer main chain, thereby giving the material good toughness and impact resistance. Both polyester and polyether monomers contain relatively long flexible segments, which can play a good role in flexibility. By regulating the ratio of rigid monomers and flexible monomers, the rigidity and flexibility of polyurea materials can be effectively balanced to obtain excellent comprehensive mechanical properties. Secondly, the present invention introduces a variety of functional additives in order to give polyurea elastomers excellent flame retardancy and insulation and other special properties. For example, adding toughening additives such as nano-SiO2 can improve the tensile strength and tensile modulus of the material. Nano-SiO2 particles can be evenly dispersed in the polymer matrix, and through interaction with the organic matrix, a strong interfacial bonding force is formed, thereby enhancing the mechanical properties of the material. At the same time, nano-SiO2 can also form a heat-resistant ceramic coating during the combustion process of the polymer, which plays a certain flame retardant role.

[0102] In addition, the present invention also adds inorganic salt flame retardants, such as Mg3N2 salt. During the combustion process, Mg3N2 salt can release inert gas (N2), producing a flame retardant effect of blocking heat transfer and diluting oxygen. At the same time, MgO formed after the decomposition of Mg3N2 also has a good flame retardant ceramic effect. Through the synergistic effect of additives such as nano-SiO2 and Mg3N2 salt, the flame retardancy of the polyurea material of the present invention can be effectively improved. At the same time, the polyurea elastomer of the present invention itself also has excellent insulation properties due to its high aromatic content.

[0103] In addition, the present invention also considers optimization based on reaction kinetics and genetic algorithm: first, in terms of reaction kinetics, the present invention establishes a complete set of reaction kinetic equations, accurately describing the kinetic behavior of the addition reaction of isocyanate groups with amino groups and hydroxyl groups and the hydrolysis side reaction. These equations not only take into account the influence of group concentration, but also introduce the influencing factors of catalyst concentration and stirring speed, so that the model is more in line with the actual reaction situation. The activation energy and pre-exponential factor are obtained by dynamic DSC experiment, and the influence coefficient is determined by isothermal reaction kinetic experiment, and an accurate rate constant calculation equation is established. Such a kinetic model can predict the reaction process under different conditions and provide theoretical guidance for formulation optimization. Secondly, in terms of structural prediction, the present invention establishes a prediction equation for cross-linking density and molecular weight. The cross-linking density prediction equation takes into account the influence of factors such as functionality, conversion rate, cross-linking agent concentration, ambient humidity and temperature, and calibrates each influence coefficient by swelling method. The molecular weight prediction equation takes into account the influence of factors such as chain transfer, chain termination and temperature, and is verified by GPC test. These two equations accurately reveal the relationship between reaction conditions and material microstructure, providing a basis for performance regulation. Third, in terms of performance prediction, the present invention establishes a mechanical property prediction equation to link the microstructure parameters of the material with the macroscopic performance. Through nonlinear fitting of a large amount of experimental data, various empirical constants are obtained, so that the model has a high prediction accuracy. Fourth, in terms of optimization algorithm, the present invention uses genetic algorithm for multi-objective optimization. By rationally designing decision variables, fitness functions and constraints, performance indicators and cost factors are considered simultaneously during the optimization process to ensure that the obtained optimization results are practical. In addition, the present invention also has innovations in process control, such as extending the reaction time and improving the degree of crosslinking through decompression and heating treatment; and ensuring the quality of the coating by strictly controlling the spraying parameters. These process measures are combined with theoretical models to form a complete technical system.

[0104] In order to better understand and implement the present invention, a specific embodiment of the present invention is provided below: In this embodiment, a high-insulation, flame-retardant spray polyurea elastomer material for power transmission line protection is prepared. The material is required to have excellent insulation performance (dielectric strength greater than 30kV / mm), good flame retardant performance (vertical burning test reaches V0 level), high mechanical properties (tensile strength greater than 35MPa, elongation at break greater than 400%) and good weather resistance. The specific implementation steps are as follows.

[0105] In step S10, the isocyanate component is first prepared, and 4,4'-diphenylmethane diisocyanate (MDI) is selected as the rigid isocyanate monomer, with a purity greater than 99.5% and an NCO content of 33.6%. A polyether isocyanate prepolymer (NCO content of 16.2%) is selected as the flexible isocyanate prepolymer, with a molecular weight of 2200. Under an environment of a temperature of 22° C. and a relative humidity of 45%, the two isocyanates are mixed in a mass ratio of 4:6, and the stirring speed is 300 r / min for 30 minutes to obtain the isocyanate component. Then prepare the resin component, select diethyltoluenediamine (purity greater than 99.0%) as the rigid polyamine monomer, 1,1,1-tri(4-hydroxyphenyl)ethane (purity greater than 98.5%) as the rigid polyol monomer, and polyether polyamine (molecular weight of 2000) as the flexible polyol monomer, mix the three monomers in a mass ratio of 3:2:5, stir at a speed of 300 r / min, stir for 30 minutes, and obtain the resin component.

[0106] In step S20, the traditional experimental optimization method is first carried out. Orthogonal experimental design L9 (34) is used to investigate the effects of four factors on product performance: the mass ratio of isocyanate component to resin component (1.05:1, 1.10:1, 1.15:1), the mass ratio of MDI to polyether isocyanate prepolymer (3:7, 4:6, 5:5), the mass ratio of rigid polyamine monomer, rigid polyol monomer to flexible polyamine monomer (2:2:6, 3:2:5, 4:2:4), and reaction temperature (25°C, 30°C, 35°C). After 9 groups of experiments, the optimal process parameters were obtained: the mass ratio of isocyanate component to resin component was 1.10:1, the mass ratio of MDI to polyether isocyanate prepolymer was 4:6, the mass ratio of the three monomers was 3:2:5, and the reaction temperature was 30°C. Under this condition, the tensile strength of the product is 36.2MPa, the elongation at break is 425%, and the dielectric strength is 31.5kV / mm, reaching the V0 flame retardant requirement. However, this method requires more experiments and it is difficult to consider the influence of all process parameters.

[0107] Therefore, this embodiment uses a method of establishing a set of equations and combining genetic algorithms for optimization. First, a set of reaction kinetic equations is established. The concentration of isocyanate groups is determined by silane titration, with an initial concentration of 2.45 mol / L; the concentration of amino groups is determined by potassium periodate titration, with an initial concentration of 1.12 mol / L; the concentration of hydroxyl groups is determined by acetylation, with an initial concentration of 0.86 mol / L; and the moisture content is determined by Karl Fischer titration to be 0.05%. The dynamic DSC method is used to test at four heating rates of 5, 10, 15, and 20 °C / min to obtain the activation energy and pre-exponential factor of each reaction: E 1 =52.6 kJ / mol, A 1 =2.35×104 L / (mol·s); E 2 =58.4 kJ / mol, A 2 =3.12×10 4 L / (mol·s); E 3 =45.8 kJ / mol, A 3 =1.86×10 4 L / (mol·s). Through isothermal reaction kinetics experiments, it was determined that the catalyst influence coefficient α=85.6L / mol and the stirring influence coefficient β=0.0023min / r.

[0108] When establishing the structure prediction equation, the crosslink density of different formulations was determined by the swelling method, ranging from 0.8 to 2.2 mol / L; the number average molecular weight was determined by GPC, ranging from 15,000 to 35,000; the flexible segment content was determined by 1H-NMR, ranging from 35% to 65%. Based on multiple sets of experimental data, the crosslink density influence coefficient γ = 42.3 L / mol, the humidity influence coefficient δ = 0.015, and the temperature influence coefficient λ = 0.028K were obtained. -1 The molecular weight influence coefficient μ = 96.5 L / mol, the chain termination influence coefficient ν = 0.45, and the temperature influence coefficient ξ = 0.021 K were obtained. -1 .

[0109] like Figure 2 As shown, the change of the conversion rate of the polyurea reaction at different temperatures over time is shown to illustrate the effect of temperature on the reaction rate. The curve shows that as the temperature increases, the reaction rate increases and the final conversion rate tends to be stable.

[0110] In the performance prediction equation, the tensile strength (range 30-40 MPa) and elongation at break (range 350%-500%) of 30 groups of samples with different formulations were measured by a universal testing machine, and nonlinear fitting was performed in combination with the structural parameters to obtain the empirical constants: K 1 =12.6, a=0.65, b=0.28, c=0.018, d=0.025, eta=6.8; K 2 =865, e=-0.42, f=0.35, g=0.032, h=0.021, κ=4.2.

[0111] Table 1 Performance test results under different test conditions.

[0112] Test No. Crosslink density (mol / L) Molecular weight Tensile strength(MPa) Elongation at break (%) 1 1.25 22000 33.5 420 2 1.45 25000 35.2 395 3 1.65 28000 36.8 380 4 1.85 31000 38.5 365 5 2.05 33000 39.2 355

[0113] like Figure 3As shown, the comprehensive effect of cross-linking density and molecular weight on tensile strength is demonstrated, which is used to intuitively demonstrate the relationship between structural parameters and performance. It can be seen from the figure that with the increase of cross-linking density and molecular weight, the tensile strength shows a nonlinear growth trend.

[0114] In the genetic algorithm optimization, the target value is set: number average molecular weight Crosslink density * =1.8mol / L, tensile strength σ * =38MPa, elongation at break ε * =400%. The weight coefficient is determined by the hierarchical analysis method: w 1 =0.20, w 2 =0.25, w 3 =0.25, w 4 =0.20, w 5 =0.10. Set component constraint: 0.3≤x 1 ≤0.7, 0.3≤x 2 ≤0.7, 0.2≤x 3 ≤0.5, 0.2≤x 4 ≤0.5, 0.3≤x 5 ≤0.5. Set process parameter constraints: 293K≤T≤333K, 100r / min≤ω≤500r / min, 0.001mol / L≤C cat ≤0.01mol / L, 0.01≤C cross ≤0.05,0.001mol / L≤C trans ≤0.01mol / L, 0.01≤C f ≤0.15, 3600s≤t≤7200s. Figure 4 As shown, the changes in fitness values ​​during the optimization process are shown, which shows the convergence of the optimization process. The fitness values ​​tend to be stable after about 150 generations.

[0115] After 300 generations of iterative optimization, the optimal solution was obtained: the mass ratio of MDI to polyether isocyanate prepolymer was 4.2:5.8, the mass ratio of rigid polyamine monomer, rigid polyol monomer and flexible polymonomer was 3.2:2.1:4.7, the reaction temperature was 303K (30°C), the stirring speed was 320r / min, the catalyst concentration was 0.005mol / L, the crosslinking agent content was 0.03, the chain transfer agent concentration was 0.004mol / L, the filler content was 0.08, and the reaction time was 5400s. Under the optimal process conditions, the subsequent steps were completed.

[0116] In step S30, the isocyanate component and the resin component are mixed in an optimized ratio, the stirring speed is controlled at 320 r / min, and the reaction temperature is controlled at 30°C.

[0117] In step S40, firstly, the silane coupling agent KH550 is dissolved in a mixed solution of ethanol and water (volume ratio 95:5), and allowed to stand for 2 hours, and the added amount is 1.2% of the polymer solid content. Nano silicon oxide (particle size 20nm) is surface modified with KH550, dried at 80°C for 4 hours, and the added amount is 3.5% of the polymer solid content. Mg3N2 salt is dried at 105°C for 2 hours, and the added amount is 8.5% of the polymer solid content. The three additives are added to the polymer solution in sequence and stirred evenly.

[0118] In step S50, the mixed solution is stirred at 30°C for 90 minutes, during which the viscosity increases from the initial 3500 mPa·s to 5800 mPa·s, the solid content increases from 95.2% to 96.8%, and the free NCO group content decreases from 11.2% to 9.5%. Figure 5 As shown, the changes in viscosity and free NCO content during the reaction are shown, showing that as the reaction proceeds, the viscosity of the system gradually increases while the free NCO content gradually decreases.

[0119] In step S60, the reaction system is decompressed to 40 mmHg, heated to 62°C, and maintained for 90 minutes. At this time, the system viscosity is 7200 mPa·s, the solid content is 98.5%, and the free NCO group content is 6.8%.

[0120] In step S70, a high-pressure airless spraying device is used, with a spray gun pressure of 0.5 MPa and a spray distance of 30 cm, to spray on the surface of the steel plate that has been cleaned with acetone and polished with 180-grit sandpaper. A cross-spraying method is adopted with an overlap rate of 55%, and a coating with a thickness of 1.5 mm is obtained.

[0121] In step S80, the sample is cured for 36 hours in an environment of 25°C and 60% relative humidity, with a surface drying time of 45 minutes and a thorough drying time of 5 hours.

[0122] In step S90, the performance of the cured material is tested: the tensile strength is 38.5 MPa, the elongation at break is 415%, the Shore hardness is 85A, the cross-linking density is 1.82 mol / L, the residual stress is 3.8 MPa, the dielectric strength is 32.5 kV / mm, and the vertical burning test reaches V0 level.

[0123] like Figure 6As shown, compared with the traditional optimization method, the method of combining the establishment of a set of equations and the genetic algorithm used in the present invention has the following advantages: 1. The optimization efficiency is significantly improved. The traditional method needs to complete at least 9 groups of experiments, while the present method can quickly obtain the optimal solution through theoretical calculation. 2. The optimization results are better. Under the same raw material system, the product performance obtained by this method (tensile strength 38.5MPa vs 36.2MPa, dielectric strength 32.5kV / mm vs 31.5kV / mm) is significantly better than that of the traditional method. 3. The optimization process is more comprehensive, and the influence of multiple process parameters can be considered at the same time, and the obtained process parameter combination is more reasonable. 4. The prediction accuracy is high. The product performance can be accurately predicted by the established set of equations, and the error between the predicted value and the measured value is within 5%.

[0124] It should be noted that the variables involved in the present invention are explained in detail as shown in Table 2 below.

[0125] Table 2 Variable explanation table

[0126]

[0127] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for preparing a highly insulating, flame-retardant spray polyurea elastomer material, characterized in that: The following steps are involved: S10, preparing an isocyanate component and a resin component, wherein the isocyanate component comprises a rigid isocyanate monomer and a flexible isocyanate prepolymer, and the resin component comprises a rigid polyamine monomer, a rigid polyol monomer and a flexible polyamine monomer; S20, using a genetic algorithm to calculate the optimal composition ratio of the isocyanate component and the polyol component according to the target molecular weight, crosslinking density and mechanical property requirements; S30, mixing the isocyanate component and the resin component according to the ratio optimized by the genetic algorithm to form a polymer solution; S40, adding an appropriate amount of additives, including coupling agent, toughening agent, and flame retardant, to the polymer solution; S50, stirring the mixed solution for a certain period of time to form an intermediate solution; S60, decompressing and heating the intermediate solution to extend the reaction time; S70, pouring or spraying the reacted solution on the surface of the sampling plate; S80, heating and curing the material of the sprayed sampling plate to form a cross-linked polyurea elastomer coating; S90, conduct physical and chemical performance tests and evaluations on the cured polyurea elastomer material, including tensile strength, elongation at break, insulation performance, and flame retardant performance, to complete the preparation of high insulation and flame retardant spray polyurea elastomer material.

2. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The rigid isocyanate monomer is 4,4'-diphenylmethane diisocyanate.

3. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The flexible isocyanate prepolymer is a polyester or polyether isocyanate prepolymer.

4. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The rigid polyamine monomer is diethyltoluenediamine.

5. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The rigid polyol monomer is 1,1,1-tris(4-hydroxyphenyl)ethane.

6. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The flexible multi-monomer is a polyether polyamine or a polyether polyol with a molecular weight greater than 1000.

7. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The toughening agent is nano silicon oxide.

8. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The flame retardant is Mg3N2 salt.

9. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: The stirring reaction time in step S50 is controlled to be 1-2 hours.

10. The method for preparing a highly insulating and flame-retardant spray polyurea elastomer material according to claim 1, characterized in that: In the step S60, a neural network model based on an encoder-decoder structure is used to control the decompression heating treatment of the intermediate solution. The neural network model includes a five-layer pyramid structure. The five-layer pyramid structure sequentially processes the pressure ranges of 200 to 250, 150 to 200, 100 to 150, 50 to 100, and 0 to 50 mmHg and predicts the free isocyanate group content in the percentage ranges of 11 to 12, 9 to 11, 8 to 9, 6 to 8, and 5 to 6, respectively. The neural network model is trained based on 300 groups of industrial production historical data.

Images