A design method of a multi-principal-element high-entropy system-based cross-scale biomimetic gradient coating

Abstract

The application belongs to the technical field of high-performance tool coating, and relates to a design method of a cross-scale biomimetic gradient coating based on a multi-principal-element high-entropy system, which sequentially forms a high-entropy alloy plastic layer, a high-entropy alloy / high-entropy ceramic strong and tough layer and a nanocrystalline high-entropy ceramic superhard layer on the surface of a tool base body, constructs a gradient transition structure, matches the high-entropy alloy plastic layer with a bonding phase of the tool base body, optimizes the high-entropy alloy / high-entropy ceramic strong and tough layer material, improves the interface bonding force, adopts thermal-mechanical coupling calculation to optimize the gradient transition design, constructs an atomic model and optimizes the microstructure based on molecular dynamics simulation, establishes a finite element model to simulate thermal cycles and mechanical loads, optimizes the interface performance, constructs a mathematical model by using a BP neural network and a particle swarm optimization algorithm, and realizes the gradual transition design of the tool base body and the coating. The application solves the problems of large brittleness, insufficient adhesion and limited thermal stability of the tool coating interface, and improves the interface bonding performance of the coating, the stability and durability of the coating tool.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of high-performance tool coating, and in particular relates to a design method of a cross-scale biomimetic gradient coating based on a multi-principal-element high-entropy system. BACKGROUND

[0002] In the modern manufacturing field, tool coating technology is widely used to improve the machining performance of tools. Mainly by depositing single or multi-layer high-performance coatings on the surface of tool substrates to enhance the wear resistance, oxidation resistance and high-temperature stability of the tools. Traditional tool coating materials, such as TiC, TiN and AlTiN, show good service life and stability in conventional machining environments. However, under high temperature, high pressure and extreme working conditions such as aerospace and mold manufacturing, these coatings are prone to failure due to thermal stress concentration, crack propagation and brittle spalling, etc., resulting in a shortened service life of the tool and affecting the machining quality.

[0003] In order to improve the bonding performance of the tool substrate and the coating, the existing technology mainly adopts physical and chemical treatment methods. Physical treatment methods include plasma etching and oxidation treatment, etc. These methods can remove the adhesive phase components on the surface layer of the blade substrate that are not conducive to the growth of the coating, thereby improving the bonding force between the coating and the substrate. However, due to the possibility of generating pores or surface defects during the treatment process, it may lead to a decrease in the overall fracture strength of the tool. Chemical treatment methods mainly use carbon, nitrogen, titanium and other element penetration technology to improve the nucleation density, quality and bonding strength of superhard coatings (such as diamond coatings) on the surface layer of the blade substrate. However, due to the low element diffusion rate, the penetration rate of this method is limited, and it is difficult to accurately control the morphology of the product after penetration, thereby affecting the final performance of the tool.

[0004] In addition, setting a transition layer is one of the important methods to improve the bonding force between the tool substrate and the coating. Common transition layers include single-layer metal transition layers such as Ti, Cr, W, and multi-component composite transition layers such as TiZrNbTa, CrNbTaN, TiCNCr. Among them, single-layer metal transition layers can promote the rapid nucleation and growth of superhard phases on the surface of the transition layer, and the process is relatively mature. However, in view of the diversified demand for difficult-to-machine materials, single-layer transition layers may not provide sufficient multifunctionality in some application scenarios, so research has gradually shifted to multi-component composite transition layers. In recent years, high-entropy alloy (HEA) materials such as TiZrNbTaMo, TiZrHfNb and TiZrHfNbTa have been applied to some hard alloy tool coatings due to their excellent adhesion, ductility and strength, and have shown good bonding force and anti-peeling ability.

[0005] Although the existing tool coating technology has made some progress, it still faces problems such as large interface brittleness, insufficient adhesion, limited thermal stability, etc. Therefore, how to further optimize the tool coating structure, improve the bonding force between the substrate and the coating, reduce the interface stress concentration, and improve the wear resistance and oxidation resistance of the coating, is still a key problem that needs to be solved in the current technical field. SUMMARY

[0006] In view of the above problems, the present application proposes a cross-scale biomimetic gradient coating design method based on a multi-principal-element high-entropy system by combining gradient coating structure design theory and comprehensively considering the tool substrate surface composition and processing integrity, the excellent adhesion and toughness of high-entropy alloy materials, and the wear resistance of high-entropy ceramic (HEC) materials. The present application solves the problems of stress concentration, cracking and brittle peeling of the existing tool coating interface due to high temperature and high pressure working conditions during cutting process, improves the interface bonding performance of the tool coating, and improves the stability and durability of the tool coating under extreme working conditions.

[0007] The present application provides a cross-scale biomimetic gradient coating design method based on a multi-principal-element high-entropy system, which comprises:

[0008] Step 1: forming a high-entropy alloy plastic layer, a high-entropy alloy / high-entropy ceramic tough layer, and a nanocrystalline high-entropy ceramic superhard layer on the tool substrate surface in sequence to construct a biomimetic gradient coating with a gradient transition structure;

[0009] Step 2: determining the material composition of each layer of the tool substrate and the gradient coating, matching the material of the high-entropy alloy plastic layer in the gradient coating with the high-entropy alloy binder phase in the tool substrate, selecting the material of the nanocrystalline high-entropy ceramic superhard layer according to the target machining object of the tool, and determining the material of the high-entropy alloy / high-entropy ceramic composite tough layer based on the materials of the high-entropy alloy plastic layer and the nanocrystalline high-entropy ceramic superhard layer;

[0010] Step 3: analyzing the interface performance of the biomimetic gradient coating using a thermal-mechanical coupling calculation method, optimizing the gradient transition design of the interface composition based on the calculation and analysis results, and determining a gradual transition scheme of the interface material composition;

[0011] Step 4: constructing an atomic model reflecting the interface structure of the high-entropy alloy plastic layer, the high-entropy alloy / high-entropy ceramic composite tough layer, and the nanocrystalline high-entropy ceramic superhard layer, performing molecular dynamics simulation on the atomic system under the set temperature, pressure, and boundary conditions, obtaining data of the lattice and grain parameters, the grain boundary width, and the phase ratio changing with the position, and determining a gradual transition scheme of the microstructure based on the obtained data, so that the layers of the coating are smoothly connected in terms of lattice, grain, grain boundary, and phase structure;

[0012] Step 5: According to the actual process parameters and the temperature-dependent mechanical properties of each layer of material, a finite element model containing the tool substrate and each layer of the gradient coating is established. In the model, a heat-force coupling load condition is applied to simulate the thermal cycle and mechanical load during the working process to obtain the distribution of stress, strain and residual stress at the interface, analyze the stress concentration and deformation in the interface region, and determine an adaptive and optimized interface mechanical property transition design scheme.

[0013] Step 6: The obtained data of the composition, microstructure and mechanical properties of each layer of material are preprocessed and feature extracted to form a multidimensional data set. The data is trained and optimized using a BP neural network and a particle swarm optimization algorithm to construct a mathematical model reflecting the interrelation of composition, structure and performance between each layer of the gradient coating. The mathematical model is embedded into a MATLAB mathematical modeling platform to realize the gradual transition design between the tool substrate and the gradient coating and dynamically control the interface parameters.

[0014] In the preferred implementation manner, further, in step 2, the material of the tool substrate includes but is not limited to any one of TiC-CrFeCoNi, Ti(C,N)-CoCrCuFeNi, TiC-CrMnFeCoNi.

[0015] In the preferred implementation manner, further, in step 2, the material of the high-entropy alloy plastic layer includes but is not limited to any one of CrFeCoNi, CoCrCuFeNi, CrMnFeCoNi; the material of the nanocrystalline high-entropy ceramic superhard layer includes but is not limited to any one of AlTiSiN, (TiZrNbTaMo)C / N, (TiAlTaCrZr)N; the material of the high-entropy alloy / high-entropy ceramic composite toughening layer includes but is not limited to any one of (CrFeCoNi)-AlTiSiN, (CoCrCuFeNi)-(TiZrNbTaMo)C.

[0016] In the preferred implementation manner, further, step 3 includes:

[0017] Step 3.1: Select a database suitable for the biomimetic gradient coating in Thermo-Calc software, establish a biomimetic gradient coating structure model, and set the required simulation environment parameters;

[0018] Step 3.2: Calculate the interface binding energy of the substrate, the high-entropy alloy plastic layer, the high-entropy alloy / high-entropy ceramic toughening layer, and the nanocrystalline high-entropy ceramic superhard layer, and adjust the composition of the interface material according to the calculation results;

[0019] Step 3.3: Calculate the interface residual stress of each layer of the biomimetic gradient coating under thermal cycle and mechanical load conditions;

[0020] Step 3.4: Calculate the difference in thermal expansion coefficients between the substrate and the plastic layer of high-entropy alloy in the biomimetic gradient coating, and between adjacent layers in the biomimetic gradient coating, and optimize the thermal matching of the interface accordingly;

[0021] Step 3.5: Calculate the phase stability between the substrate and the biomimetic gradient coating under high temperature environment.

[0022] In a preferred implementation, further, in step 2, the material of the tool substrate comprises a composite of high-entropy alloy binder phase and ceramic hard phase, wherein the high-entropy alloy binder phase in the tool substrate material is the same as the high-entropy alloy binder phase in the material of the plastic layer of high-entropy alloy.

[0023] In a preferred implementation, further, step 5 comprises:

[0024] Step 5.1: Establish a three-dimensional geometric model containing the tool substrate and each layer of the gradient coating, specially divide the interface area to form independent interface elements or use adhesive contact, and clearly define the transition area between different layers;

[0025] Step 5.2: Assign temperature-dependent elastic modulus, Poisson's ratio, thermal expansion coefficient, yield strength and other plasticity parameters to each layer respectively;

[0026] Step 5.3: Reasonably divide the grid for the entire model, use local encryption grid in the interface and areas with high stress gradient, and determine the optimal grid size through grid independence verification;

[0027] Step 5.4: According to the actual process flow, apply thermal load and simulate the thermal cycle process to form a temperature gradient;

[0028] Step 5.5: Apply fixed boundary conditions or appropriate constraints to the tool substrate, apply external force or displacement load in the working state to the coating and interface area, simulate the mechanical response generated in the machining or use process, and set the interface contact properties at the same time;

[0029] Step 5.6: Use a nonlinear thermal-force coupled solver to jointly solve the temperature field and stress field, and ensure convergence through step-by-step solving;

[0030] Step 5.7: Extract stress, strain, residual stress and local deformation data in the key area, use post-processing tools to generate temperature field, stress field distribution map and local enlarged view, and clearly define the stress concentration position and amplitude;

[0031] Step 5.8: Compare the maximum stress, strain distribution and residual stress level of the interface area under different design schemes, establish a stress concentration index, use the target function as the basis for optimization, and analyze the influence of different interface contact parameters, layer thickness and material transition zone design on the target function;

[0032] Step 5.9: Sensitivity analysis is performed on the key parameters to determine their impact on the mechanical response. Based on the results of the sensitivity analysis, design improvement schemes are proposed to optimize the interface transition design.

[0033] Step 5.10: The optimized design scheme is simulated under multiple working conditions. The verification results are fed back to the model for multiple iterations until an interface mechanical performance transition scheme that meets the actual working condition requirements is obtained.

[0034] In the preferred implementation, further, step 6 includes:

[0035] Step 6.1: Collect the multi-scale data obtained in steps 3-5, and perform missing value elimination, outlier processing, and normalization conversion on the original data. Use principal component analysis or feature extraction techniques to screen key indicators related to interface bonding strength, toughness, and overall performance, construct an input data matrix, and use it as the basis data for subsequent modeling.

[0036] Step 6.2: Define component entropy, structure entropy, and performance entropy, and construct a unified index system with the entropy values of component entropy, structure entropy, and performance entropy. Build a multi-objective optimization model and set the collaborative control target.

[0037] Step 6.3: Select a multi-layer feedforward neural network as the basic model structure. The input layer receives the preprocessed component, structure, and performance data, and the output layer outputs the interface bonding strength and toughness prediction values. Set the number of hidden layers and nodes according to the data complexity, and use activation functions to improve the model's nonlinear mapping ability.

[0038] Step 6.4: Train the network using the error backpropagation algorithm, set the loss function to mean square error or other appropriate indicators, use cross-validation to evaluate the model's generalization ability, adjust the network parameters until the training error and validation error meet the expected requirements, use the MATLAB mathematical modeling platform to implement the training process, and generate a training report and performance evaluation chart.

[0039] Step 6.5: Use the PSO algorithm to perform global search on the weights, biases, and hyperparameters of the BP neural network to further reduce the training error.

[0040] Step 6.6: Integrate the "component entropy-structure entropy-performance entropy" collaborative control model that has undergone data preprocessing, BP neural network training, and PSO optimization into the MATLAB platform to establish a systematic simulation calculation framework. Use historical data and newly generated simulation data to verify the model, make the model prediction results consistent with the actual performance indicators, analyze the prediction error and sensitivity, and feed back adjustments to the feature engineering, network structure, and optimization parameters to form a closed-loop improvement process.

[0041] In the preferred implementation, further, in step 4, the process of generating the optimal microstructure gradient transition scheme includes: first, according to the crystal structure, lattice parameters and defect information of the selected high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic composite toughening layer and nanocrystalline high-entropy ceramic superhard layer, an atomic model meeting the periodic and non-periodic boundary conditions is constructed, and the atomic arrangement and transition zone of the interface region are designed; then, according to the characteristics of each material system, the applicable interatomic potential is selected, the initial atomic position and velocity are set, and the energy minimization method is used to obtain the system equilibrium configuration, and then the Nosé-Hoover temperature controller is used to control the constant temperature or temperature gradient under the simulated real heat treatment conditions, and the interface atomic rearrangement, grain boundary diffusion and local phase change phenomena are captured by refining the time step and multi-step simulation; finally, the radial distribution function, coordination number statistics and local strain analysis methods are used to collect and quantitatively evaluate the simulation data, the interface energy, bonding strength and lattice matching degree are calculated, and the optimal microstructure gradient transition scheme is determined.

[0042] In the preferred implementation, further, in step 6.5, the process of globally searching the weights, biases and hyperparameters of the BP neural network through the PSO algorithm includes: initializing a certain number of particles, defining the fitness function as the prediction error of the network on the validation set, updating the speed and position of each particle, iteratively searching until convergence, finally obtaining the optimal parameter set, and applying it to the BP neural network model to form the comprehensive optimized synergistic control model.

[0043] In the preferred implementation, further, in step 6.2, the composition entropy is used to reflect the uniformity and diversity of the composition and element distribution of each layer of material, and the calculation formula can refer to the information entropy model; the structure entropy is used to describe the order degree of lattice matching, grain boundary width and defect distribution in the interface microstructure; the performance entropy is used to quantify the balance and robustness of the interface mechanical properties.

[0044] The beneficial effects of the present application are:

[0045] Firstly, the method of the application makes full use of the unique advantages of multi-principal-element high-entropy system, constructs a biomimetic gradient coating composed of a high-entropy alloy plastic layer, a high-entropy alloy / high-entropy ceramic strong and tough layer, and a nanocrystalline high-entropy ceramic superhard layer, realizes precise cross-scale design from atomic scale to engineering scale, and effectively solves the problems of thermal stress concentration, crack propagation and brittle peeling of traditional coating under high temperature, high pressure and extreme working conditions. Compared with the single material or simple multi-layer structure (such as TiC, TiN, AlTiN coating) in traditional technology, which is prone to tool failure due to large interface brittleness and insufficient adhesion, the existing physical and chemical treatment and single or composite transition layer can improve the bonding force between the substrate and the coating to some extent, but it is still difficult to fully meet the performance requirements under high load conditions. The application is optimized and designed to overcome these defects. Specifically, by using various numerical simulation techniques such as thermal-mechanical coupling calculation, molecular dynamics simulation and finite element analysis, the smooth transition of each layer of the coating in lattice, grain boundary and phase structure can be accurately controlled, thereby effectively relieving the problems of interface stress concentration and local deformation. At the same time, by introducing BP neural network and particle swarm optimization algorithm for intelligent analysis and dynamic regulation of multi-dimensional data, the interface parameters are optimized in real time, and the bonding strength, interface toughness, thermal matching and fatigue resistance between the coating and the tool substrate are improved, which not only prolongs the service life of the coated tool, but also improves the machining efficiency, meeting the demand for high-performance tools in modern manufacturing field.

[0046] Secondly, in the preferred implementation mode, the material combination of the tool substrate, the high-entropy alloy plastic layer, the nanocrystalline high-entropy ceramic superhard layer and the high-entropy alloy / high-entropy ceramic composite strong and tough layer realizes the balance of high hardness, high wear resistance and good toughness through the optimization of the synergistic effect of the tool substrate and the multi-layer composite structure.

[0047] Thirdly, in the preferred implementation mode, step 3 of the application optimizes and adjusts the interface bonding energy, residual stress, thermal expansion matching and phase stability at high temperature between each layer of the coating through accurate simulation, realizes the optimal configuration of material components and interface structure, thereby improving the overall heat resistance, wear resistance and fatigue resistance of the coating, prolonging the service life of the substrate and the coating, and ensuring that the system can work stably and reliably under complex thermal cycling and mechanical load conditions.

[0048] Fourthly, in the preferred implementation mode, the high-entropy alloy binder phase used in the tool substrate and the plastic layer is completely consistent, thereby realizing the best matching of chemical and mechanical properties at the material interface. This matching not only enhances the interface bonding force, but also reduces the internal stress caused by thermal expansion mismatch.

[0049] Fifth, in the preferred implementation, step 5 of the present invention achieves accurate simulation of the temperature gradient and stress distribution at the interface between the tool substrate and the gradient coating by meticulously establishing a three-dimensional geometric model, assigning temperature-dependent material parameters, and refining the local mesh. It can accurately capture the stress concentration area and residual stress level. By using a nonlinear thermo-mechanical coupling solver and multi-condition iterative verification, it not only establishes objective evaluation indicators for different design schemes, but also optimizes the interface transition design through sensitivity analysis, thereby improving the heat resistance, wear resistance and fatigue resistance of the tool.

[0050] Sixth, in the preferred implementation, step 6 of the present invention achieves effective integration of multi-scale data and fine extraction of key features through the model, constructs a unified index system of "composition entropy – structural entropy – performance entropy", and successfully maps the complex relationship between interface bonding strength, toughness and overall performance by means of multi-objective optimization, BP neural network and PSO global search. Finally, the closed-loop simulation and feedback improvement mechanism not only significantly improves the accuracy of prediction and the generalization ability of the model, but also enhances the robustness to abnormal data.

[0051] Seventh, in the preferred implementation, step 4 of the present invention can accurately capture atomic rearrangement, grain boundary diffusion and local phase transformation in the interface region during heat treatment, and achieve optimized matching of crystal structure and interface characteristics, thereby improving the interface bonding strength and overall lattice matching degree, and reducing the internal stress and defect diffusion risk of the coating. Attached Figure Description

[0052] Figure 1 This is a flowchart of a cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system, according to an embodiment of the present invention.

[0053] Figure 2 This is a schematic diagram of the high-entropy alloy / ceramic cutting tool matrix material structure according to an embodiment of the present invention;

[0054] Figure 3 This is a high-entropy multi-principal gradient coating architecture diagram according to an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of the grain and grain boundary microstructure of an embodiment of the present invention.

[0056] Among them, 1-high-entropy alloy binder phase; 2-ceramic hard phase; 3-high-entropy alloy plastic phase; 4-high-entropy ceramic superhard phase; 5-tool matrix; 6-high-entropy alloy plastic layer; 7-high-entropy alloy / high-entropy ceramic toughening layer; 8-nanocrystalline high-entropy ceramic superhard layer; 9-grain; 10-grain boundary. Detailed Implementation

[0057] In order for those skilled in the art to better understand the technical solutions of the present application, the present application will be further described in detail below with reference to the drawings and embodiments.

[0058] The up, down, left, right, front and back orientation terms in the present application are established based on the positional relationship shown in the drawings. If the drawings are different, the corresponding positional relationship may also change accordingly, and therefore cannot be understood as a limitation on the scope of protection.

[0059] In the present application, the terms "mounting", "connection", "interface", "connection", "fixing" and the like should be understood broadly, for example, it can be fixed connection, it can also be detachable connection, it can also be integrally connected, it can also be mechanical connection, it can also be electrical connection or can communicate with each other, it can also be direct connection, it can also be indirect connection through an intermediate medium, it can be the interconnection of two components, or it can be the interaction relationship between two components. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0060] The present application is inspired by the material composition, microstructure scale and mechanical property gradient in the shell layer structure of the biological kingdom, and combines the excellent adhesion of high-entropy alloy (HEA), the excellent wear resistance of high-entropy ceramic (HEC) and the design theory of gradient coating, to propose a cross-scale biomimetic gradient coating design method based on multi-principal high-entropy system, aiming to optimize the interface bonding performance of the tool coating and improve its stability and durability under extreme working conditions. The method constructs a gradient coating structure composed of HEA plastic layer, HEA / HEC tough layer and nanocrystalline HEC superhard layer, selects HEA plastic phase materials matching the tool substrate, optimizes the interface material transition through thermal-mechanical coupling calculation, molecular dynamics simulation and finite element analysis, uses BP neural network, particle swarm optimization algorithm (PSO) and MATLAB mathematical modeling to construct a "composition entropy-structure entropy-performance entropy" collaborative optimization model, and comprehensively improves the interface performance of the coating. At the same time, through the biomimetic gradient design, the composition and microstructure are adjusted layer by layer to realize the best matching of the toughness of the coating and the superhard property of the surface layer, thereby reducing the interface brittleness, improving the adhesion, enhancing the wear resistance and impact resistance, and finally overcoming the bottleneck problem of large interface brittleness and insufficient adhesion of the tool coating in the prior art, and improving the performance stability and service life of the tool under extreme working conditions such as high temperature, high speed and high load cutting.

[0061] Referring to the description of the drawings Figure 1 A cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system, the method comprising:

[0062] Step 1: Forming high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic strong and tough layer, and nanocrystalline high-entropy ceramic superhard layer on the surface of the tool substrate in sequence to construct a biomimetic gradient coating with gradient transition structure.

[0063] The purpose of step 1 is to construct the overall architecture of the biomimetic gradient coating. By reasonably designing the material composition and structural layout, the interface bonding force is optimized, and the wear resistance and impact resistance of the tool are improved. This design takes into account high strength, high toughness and high wear resistance, so that the tool has better performance and longer service life in high-speed, high-temperature and high-load cutting environments.

[0064] Referring to the drawings accompanying the specification Figures 2-3 , the present application provides a high-performance gradient coating tool. The tool includes a tool substrate 5 composed of a high-entropy alloy binder phase 1 and a ceramic hard phase 2. The substrate combines the good ductility of high-entropy alloy with the high hardness of ceramic hard phase, ensuring that the tool has excellent strength-toughness synergistic optimization characteristics in high-load cutting environments, improving overall service life and cutting performance.

[0065] In the interface region between the tool substrate 5 and the coating, a high-entropy alloy plastic layer 6 is provided. The high-entropy alloy plastic layer 6 is composed of a high-entropy alloy plastic phase 3 (abbreviated as HEA plastic phase, High-Entropy Alloy). The high-entropy alloy material in the plastic phase has excellent bonding performance and high ductility, which can effectively improve the bonding strength of the substrate and the coating, and at the same time, relieve the impact stress caused by cutting force, to reduce interface peeling and micro-crack propagation, and improve the durability and stability of the coating as a whole.

[0066] In the outermost layer of the coating, a nanocrystalline high-entropy ceramic superhard layer 8 is deposited. The nanocrystalline high-entropy ceramic superhard layer 8 is composed of a high-entropy ceramic superhard phase 4 (abbreviated as HEC superhard phase, High-Entropy Ceramic). The nanocrystalline high-entropy ceramic material has ultra-high hardness, excellent wear resistance and excellent high-temperature oxidation resistance, ensuring that the tool can maintain stable cutting performance in high-speed cutting and high-temperature environments, and significantly improving the wear resistance of the tool.

[0067] Between the high-entropy alloy plastic layer 6 and the nanocrystalline high-entropy ceramic superhard layer 8, a high-entropy alloy / high-entropy ceramic strong and tough layer 7 is provided. The high-entropy alloy / high-entropy ceramic strong and tough layer 7 is composed of HEA strong and tough phase and HEC strong and tough phase, forming a composite gradient structure of high-entropy alloy and high-entropy ceramic. This strong and tough layer not only ensures the gradual change of the composition of the coating from the substrate to the surface, effectively reducing the interface stress, but also enhances the interlayer bonding strength, realizing the synergistic enhancement of the overall strength and toughness. In addition, the HEA strong and tough phase provides good impact resistance, while the HEC strong and tough phase improves wear resistance and high-temperature stability, making the coating have excellent fatigue resistance and long-term stable mechanical properties.

[0068] It should be noted that the differences between the plastic phase, the tough phase and the superhard phase in step 1 in material science mainly reflect in hardness, toughness and structural function, and they play different roles in the design of gradient coating. The plastic phase has good plastic deformation ability, can enhance the bonding force between the coating and the substrate, adapt to the changes of thermal stress and mechanical stress in the cutting process, can improve the bonding force of the coating and buffer stress. The tough phase has certain hardness and toughness, connects the plastic phase and the superhard phase, forms a smooth gradient transition, reduces the interfacial stress between different phases, and improves the overall mechanical properties of the coating. The superhard phase is composed of high hardness and wear-resistant materials, which can reduce the wear of the tool in high-speed cutting and improve the wear resistance and corrosion resistance of the tool surface.

[0069] Step 2: Determine the material composition of the tool substrate and each layer of the gradient coating, so that the material of the high-entropy alloy plastic layer in the gradient coating matches the high-entropy alloy binder phase in the tool substrate, select the material of the nanocrystalline high-entropy ceramic superhard layer according to the target machining object of the tool, and determine the material of the high-entropy alloy / high-entropy ceramic composite tough layer based on the material of the high-entropy alloy plastic layer and the material of the nanocrystalline high-entropy ceramic superhard layer.

[0070] The purpose of step 2 is to reasonably select and optimize the material components of the gradient coating to ensure the bonding strength between the tool substrate and the coating, and to improve the wear resistance, impact resistance and high temperature stability of the tool.

[0071] Specifically, the tool substrate material needs to have high strength, high toughness and good wear resistance, and at the same time can provide good bonding ability with the coating. In this application, the material of the tool substrate includes but is not limited to any one of TiC-CrFeCoNi (titanium carbide-chromium iron cobalt nickel high-entropy alloy), Ti(C,N)-CoCrCuFeNi (titanium carbonitride-cobalt chromium copper iron nickel high-entropy alloy), TiC-CrMnFeCoNi (titanium carbide-chromium manganese iron cobalt nickel high-entropy alloy). If the machining object is titanium alloy, stainless steel and other difficult-to-machine materials, TiC-CrFeCoNi can be selected, which has high thermal cracking resistance and toughness. If the machining object is high hardness steel (HRC 50 or more) or heat-resistant alloy, Ti(C,N)-CoCrCuFeNi can be selected, which has good wear resistance and oxidation resistance. If the machining object is high-strength composite material, carbon fiber reinforced composite material (CFRP) and the like, TiC-CrMnFeCoNi can be selected, which can provide better thermal stability and fracture toughness.

[0072] The selection of the HEA (high-entropy alloy) plastic phase material needs to be consistent with the high-entropy alloy binder phase material of the tool substrate to improve the interface bonding strength, reduce the interface stress concentration, and improve the impact resistance. In the present application, the material of the high-entropy alloy plastic layer includes but is not limited to any one of CrFeCoNi (chromium-iron-cobalt-nickel high-entropy alloy), CoCrCuFeNi (cobalt-chromium-copper-iron-nickel high-entropy alloy), and CrMnFeCoNi (chromium-manganese-iron-cobalt-nickel high-entropy alloy). The matching principle of the material of the high-entropy alloy plastic layer in the gradient coating with the high-entropy alloy binder phase in the tool substrate includes: 1. Ensuring that the plastic phase material and the binder phase element of the tool substrate material are the same to reduce the interface stress and compatibility problems. 2. The plastic phase should also have good ductility and impact resistance to avoid coating peeling and improve overall durability. 3. Optimizing the layer thickness of the plastic phase (usually between 1-5 μm) to ensure that it can absorb impact without over-softening to affect wear resistance. For example: if the substrate material is TiC-CrFeCoNi, the HEA plastic phase material is selected as CrFeCoNi. If the substrate material is Ti(C, N)-CoCrCuFeNi, the HEA plastic phase material is selected as CoCrCuFeNi. If the substrate material is TiC-CrMnFeCoNi, the HEA plastic phase material is selected as CrMnFeCoNi.

[0073] The HEC (high-entropy ceramic) superhard phase material determines the wear resistance, high-temperature stability, and oxidation resistance of the tool. Its selection is based on the machining object of the coated tool. In the present application, the material of the nanocrystalline high-entropy ceramic superhard layer includes but is not limited to any one of AlTiSiN (aluminum-titanium-silicon-nitride), (TiZrNbTaMo)C / N (titanium-zirconium-niobium-tantalum-molybdenum carbide / nitride), and (TiAlTaCrZr)N (titanium-aluminum-tantalum-chromium-zirconium nitride). For example: if the machining object is a high-temperature alloy, heat-resistant steel, and stainless steel (such as Inconel, Hastelloy), AlTiSiN material is selected because AlTiSiN has excellent oxidation resistance, hot hardness, and high-temperature wear resistance and can remain stable at high temperatures of 900-1000℃. If the machining object is high-strength alloy steel (HRC 50 or above) and quenched steel, (TiZrNbTaMo)C / N material is selected because this material belongs to high-entropy carbide / nitride and has ultra-high hardness (≥3500HV), excellent heat resistance, and thermal cracking resistance, making it suitable for high-speed cutting of high-hardness materials. If the machining object is an aviation composite material (such as CFRP, glass fiber reinforced plastic GFRP), (TiAlTaCrZr)N material is selected because TiAlTaCrZrN has low friction coefficient, high hardness, and high oxidation resistance, which can reduce friction heat when cutting composite materials and prevent tool adhesion and failure.

[0074] To ensure the toughness and bonding strength of the coating, the interlayer transition needs to be reasonably designed. The HEA / HEC composite layer needs to consider the compatibility of both, and avoid interface material mismatch leading to peeling or crack propagation. Not only the high hardness (close to HEC) needs to be maintained, but also the toughness (close to HEA) needs to be maintained. The overall fatigue resistance is improved through microstructure optimization (such as nanostructure, gradient transition). The composite layer material must contain high-entropy alloy (HEA) and high-entropy ceramic (HEC) two types of components, and high-entropy ceramic (HEC) nanoparticles are introduced into high-entropy alloy (HEA) to form a metal-ceramic composite phase. Therefore, in this application, the materials of the high-entropy alloy / high-entropy ceramic composite toughening layer include but are not limited to any one of (CrFeCoNi)-AlTiSiN (chromium-iron-cobalt-nickel high-entropy alloy matrix-aluminum-titanium-silicon-nitride coating) and (CoCrCuFeNi)-(TiZrNbTaMo)C (cobalt-chromium-copper-iron-nickel high-entropy alloy matrix-titanium-zirconium-niobium-tantalum-molybdenum carbide coating). The (CrFeCoNi)-AlTiSiN material is suitable for high-temperature resistance and oxidation resistance requirements, and the (CoCrCuFeNi)-(TiZrNbTaMo)C is suitable for high wear resistance working conditions.

[0075] In the material of the high-entropy alloy / high-entropy ceramic composite toughening layer, the change of the proportion of the ceramic phase (HEC) mainly affects the balance of the strength and toughness of the material, the wear resistance, the heat resistance and the impact resistance. According to different processing conditions and use environments, the proportion of the ceramic phase can be divided into low proportion, medium proportion and high proportion. For example: low ceramic phase proportion (HEC content 30%, HEA content 70%) is suitable for working conditions with large impact load (such as intermittent cutting, high-speed milling), heavy load cutting (such as machining high-strength steel, cast iron), high toughness demand (such as machining fragile materials ceramic, glass fiber reinforced composite materials). Medium ceramic phase proportion (HEC content 50%, HEA content 50%) is suitable for environments that need to consider wear resistance and toughness (such as working conditions combining continuous cutting and intermittent cutting), medium load machining (such as titanium alloy, stainless steel, quenched steel), and applications with high tool life requirements but still have certain toughness requirements. High ceramic phase proportion (HEC content 70%, HEA content 30%) is suitable for machining of super-hard workpieces (such as quenched steel with HRC>50, cemented carbide), high-speed cutting, high-temperature cutting (>1000℃), and scenarios requiring extremely low wear rate (such as ultra-precision machining).

[0076] Step 3: The interface performance of the biomimetic gradient coating is analyzed by using a thermal-mechanical coupling calculation method. According to the calculation and analysis results, the gradient transition design of the interface components is optimized, and the gradual transition scheme of the interface material components is determined.

[0077] The purpose of step 3 is to analyze the interface performance of the biomimetic gradient coating using Thermo-Calc software for thermal-mechanical coupling calculations, to evaluate the bonding strength, residual stress distribution, thermal expansion matching, and phase stability of the coating interface, to optimize the bonding, toughness, and coefficient of thermal expansion (CTE) matching, and to determine the optimal material composition and thickness distribution scheme to ensure the bonding strength, thermodynamic stability, and service environment adaptability between the coating and the substrate, thereby improving the durability and failure resistance of the coating.

[0078] It should be noted that Thermo-Calc is a thermodynamic calculation and phase diagram simulation software widely used in the fields of materials science, metallurgical engineering, and chemical engineering. It is based on the CALPHAD (Calculation of Phase Diagrams) method and can be used to calculate the phase stability, thermodynamic properties, diffusion behavior, and kinetic characteristics of multi-component alloy systems. In the design of high-entropy alloy (HEA) / high-entropy ceramic (HEC) coatings, Thermo-Calc can be used to optimize the substrate-coating interface phase matching, ensure strong bonding, and avoid interfacial brittle phases; calculate the coefficient of thermal expansion (CTE), optimize the gradient transition layer, and prevent cracks caused by temperature changes; predict high-temperature stability to ensure that high-entropy alloys / high-entropy ceramics do not decompose or form brittle phases at high temperatures; and calculate diffusion behavior using DICTRA (Diffusion-Controlled Transformations, a calculation module in Thermo-Calc software for simulating diffusion-controlled phase transformation processes) to optimize coating thickness and ensure smooth gradient transition.

[0079] Gradient layer transition refers to the gradual change of composition, microstructure, or performance parameters between different materials in a multi-layer coating or composite structure, rather than forming a sudden interface, to reduce interfacial stress, optimize bonding strength, and improve overall stability. In the design of biomimetic gradient coatings, the role of the gradient transition layer is to alleviate the performance mismatch between the substrate and the HEA plastic layer, the HEA / HEC tough layer and the HEA plastic layer, and the nanocrystalline HEC superhard layer, thereby optimizing the interfacial bonding energy, thermal expansion matching (CTE), and residual stress distribution.

[0080] Specifically, step 3 includes:

[0081] Step 3.1: Select the database suitable for biomimetic gradient coatings in Thermo-Calc software, establish the structure model of the biomimetic gradient coating, and set the required simulation environment parameters.

[0082] The database suitable for the biomimetic gradient coating includes a high-entropy alloy database (TCHEA), a ceramic database (TCTI / TCNI), an interface energy database (TCINTERFACE), and a mechanical property database (TCSMECH). Among them, the high-entropy alloy database is used to calculate the thermodynamic properties and phase stability of the HEA plastic layer. The ceramic database is used to calculate the structural stability of the HEC tough layer and the nanocrystalline superhard layer. The interface energy database is used to calculate the binding energy between each layer of the coating and the substrate, and to optimize the interfacial adhesion. The mechanical property database is used to provide residual stress and thermal expansion coefficient (CTE) data of the gradient material.

[0083] The biomimetic gradient coating structure model includes a tool substrate of high-entropy alloy material, and from the inside to the outside, an HEA plastic layer, an HEA / HEC tough layer, and a nanocrystalline HEC superhard layer.

[0084] The environmental parameters include temperature range and loading conditions. According to the working temperature range of the tool, the temperature range is defined as 25-1000°C. The loading conditions include thermal loading, isostatic pressure environment, and mechanical impact, wherein the thermal loading calculates the CTE matching of each layer, the isostatic pressure environment simulates the residual stress, and the mechanical impact evaluates the actual cutting stress.

[0085] Step 3.2: Calculate the interface binding energy of the substrate, the HEA plastic layer, the HEA / HEC tough layer, and the nanocrystalline HEC superhard layer, and adjust the composition of the interface material according to the calculation results.

[0086] Specifically, CALPHAD+ interface binding energy calculation is used to evaluate the binding energy W ad between the connected layers. The calculation targets include calculating the substrate-HEA plastic layer binding energy, calculating the HEA plastic layer-HEA / HEC tough layer binding energy, and calculating the HEA / HEC tough layer-nanocrystalline HEC superhard layer binding energy.

[0087] Taking the calculation process of the substrate-HEA plastic layer binding energy W ad as an example:

[0088] W ad = E interface -(E bulk1 +E bulk2 )(1)

[0089] In the formula, E interface represents the total energy of the substrate and the HEA plastic layer interface; E bulk1 represents the bulk energy of the substrate; and E bulk2 represents the bulk energy of the HEA plastic layer.

[0090] Firstly, the material parameters of the matrix and the HEA plastic layer are determined. In Thermo-Calc, the high-entropy alloy database (TCHEA) and the carbide database (TCTI) are selected, the matrix material is TiC-CrFeCoNi, TiC is selected as the hard phase to provide high hardness and wear resistance, and CrFeCoNi is selected as the binder phase to provide good strength and toughness. In Thermo-Calc, the TCHEA+TCNI material database is selected, and the HEA plastic layer material is CoCrFeNiTi, which has an FCC (face-centered cubic crystal structure).

[0091] Then, the total energy E interface of the interface between the matrix and the HEA plastic layer is calculated. unrelaxed interface The VASP (Vienna Ab-initio Simulation Package) or Quantum ESPRESSO is used for calculation, and the interface supercell is constructed: the matrix part, TiC (ceramic phase) selects (100) FCC crystal surface, and CrFeCoNi (metal binder phase) selects (111) FCC crystal surface. The HEA plastic layer part selects (100) FCC crystal surface for matching. The binding energy under different interface mismatches is calculated, and the original interface energy E relaxed interface is calculated first. unrelaxed interface Then, the structure relaxation calculation (Relaxation Calculation) is used to calculate the optimized interface energy E relaxation energy . Further calculation of the interface binding energy E interface = E relaxed interface / A, where A represents the interface area.

[0092] Then, the bulk energy E bulk1 and E bulk2 of the matrix and the HEA plastic layer are calculated. CoCrFeNi The first-principles calculation (DFT) is used to calculate the bulk energy of the TiC-CrFeCoNi composite structure, the (100) TiC crystal surface is selected to calculate the carbide formation energy, and then the total energy E TiC of the FCC structure CoCrFeNi alloy is calculated. bulk1 E CoCrFeNi +E TiC . The DFT is used to calculate the bulk energy E bulk2 of CoCrFeNiTi (HEA plastic layer).

[0093] The calculated E interface , Ebulk1 E bulk2 Substitute into formula (1) to calculate the bonding energy W of the matrix-HEA plastic layer. ad If W ad Negative value (W) ad <0) indicates strong interfacial bonding and a stable coating. If W ad Positive value (W) ad >0) indicates that the interface integration is weak and the interface components need to be optimized.

[0094] Furthermore, the interface composition is optimized by adjusting the alloying elements of the HEA plastic layer. In this example, the selected matrix material is TiC-CrFeCoNi, and the HEA plastic layer material is CoCrFeNiTi. Optimization can be achieved through the following methods: 1. Adding Mo or Nb to the HEA plastic layer material. Mo and Nb can form transition phases (such as NbC and MoC) with C in the matrix material, improving the metal-ceramic bonding ability of the TiC-HEA interface. 2. Reducing the Ti content in the HEA plastic layer material, as excessive Ti may lead to interface embrittlement. 3. Optimizing the Co / Cr / Ni / Fe / Ti ratio in the HEA plastic layer to make it compatible with the crystal structure of the matrix.

[0095] It's important to note that FCC (Face-Centered Cubic) and BCC (Body-Centered Cubic) are two common crystal structures for metals and alloys. These structures describe the arrangement of atoms in a crystal lattice and significantly influence the physical, mechanical, and thermodynamic properties of the material. An FCC-FCC structure refers to an interface where both materials on either side of the interface use an FCC lattice (e.g., Ni, Co, Fe, and Al primarily exist in an FCC structure). A BCC-BCC structure refers to an interface where both materials on either side of the interface use a BCC lattice (e.g., Nb, Mo, and V primarily exist in a BCC structure). When two materials have the same lattice structure at their interface (FCC-FCC or BCC-BCC), their interface matching is higher, the binding energy is usually lower, and the adhesion is stronger. If the two materials have different lattice structures (e.g., an FCC-BCC structure), the lattice mismatch is greater, the interfacial binding energy is usually higher, and residual stress and microcracks are more likely to occur.

[0096] Step 3.3: Under thermal cycling and mechanical load conditions, calculate the interfacial residual stress of each layer of the biomimetic gradient coating.

[0097] The residual stress distribution of each layer under thermal cycling and mechanical load is calculated by step 3.3 to identify the stress concentration area. The influence of the interface layer thickness change on the residual stress is evaluated to optimize the coating thickness and transition layer design to avoid stress concentration. The interface stress distribution of the coating material is optimized to avoid structural failure or new phase (such as Laves phase) precipitation caused by excessive stress. If the stress is too large, the high-temperature phase stability needs to be calculated in step 3.5, and the element content in the coating material needs to be adjusted.

[0098] Specifically, step 3.3 includes:

[0099] Step 3.3.1: Use a finite element software platform to perform multi-physical field thermal-mechanical coupling simulation on the sample to calculate the thermal stress distribution under different temperature conditions and the residual stress under different coating thickness conditions.

[0100] The finite element software platform includes Abaqus, ANSYS, etc. Through simulation, the stress response information of the coating interface under thermal load and mechanical load can be obtained.

[0101] The material model of the multi-physical field thermal-mechanical coupling simulation includes: the matrix material is TiC-CrFeCoNi, the HEA plastic layer material is CoCrFeNiTi, the HEA / HEC strong and tough layer material is CoCrFeNi-TiC / NbC, and the nanocrystalline HEC superhard layer material is Al2O3-ZrC-TiC.

[0102] Set the simulation parameters of thermal load and mechanical load. Thermal load: 25℃→1000℃→25℃, cycle 100 times. Mechanical load: simulate the impact force and shear force of the tool in the cutting process.

[0103] The thermal stress distribution under different temperatures is calculated by the following formula:

[0104] σ thermal = E x a x AT (2)

[0105] Where: E represents Young's modulus (different values for different layers); a represents the thermal expansion coefficient (CTE); and AT represents the temperature change.

[0106] The residual stress under different coating thicknesses (5-20μm) is calculated by the following formula:

[0107]

[0108] Where: σ interface represents the interface stress; v represents the Poisson's ratio; E1 and E2 represent the Young's modulus of adjacent coatings; and a1 and a2 represent the thermal expansion coefficients of adjacent coatings.

[0109] Step 3.3.2: Based on the simulation results of the thermal-mechanical coupling, the optimal coating thickness is selected according to the influence of different layer thicknesses on residual stress. The interface bonding ability is improved by optimizing the content of each component element in the interface region, and the gradient layer transition scheme is optimized by calculating the interface mismatch degree, so that the lattice matching degree of the gradient layer reaches or exceeds 95%.

[0110] The thickness of different layers is for example 5 μm, 10 μm, 15 μm. According to the influence of these thicknesses on residual stress, the coating with the minimum thickness is selected as the optimal thickness. The interface bonding ability is optimized by reducing the Ti content (reducing the interface brittleness) or increasing the Nb content (increasing the flexibility), etc.

[0111] Step 3.4: Calculate the thermal expansion coefficient (CTE) difference between the substrate and the HEA plastic layer in the biomimetic gradient coating, and between adjacent layers in the biomimetic gradient coating, and optimize the interface thermal matching accordingly.

[0112] Specifically, TCHEA (high-entropy alloy database) + TCTI (carbide database) are selected to calculate the CTE at different temperatures using Thermo-Calc:

[0113]

[0114] In the formula, α(T) represents the linear thermal expansion coefficient of the material at temperature T, with the unit of 1 / K; L represents the initial length of the material (usually measured at room temperature); dL represents the length change of the material during temperature change dT; dT represents the temperature change.

[0115] The CTE change curve of each layer material in the range of 25℃-1000℃ is calculated.

[0116] Taking the substrate and the HEA plastic layer as an example, the CTE matching degree of the substrate-TiC-CrFeCoNi and the HEA plastic layer is calculated:

[0117] ΔCTE = |CTE 基体 -CTE HEA塑性层 |(5)

[0118] If ΔCTE < 4 × 10 -6 / K, it means that the risk of interface cracking caused by thermal cycling can be reduced.

[0119] If ΔCTE ≥ 4 × 10 -6 / K, the Ti / Zr / Nb content in the HEA plastic layer is increased to reduce the CTE of the coating and match it with the substrate. Or a Ti-Zr transition layer (CTE ≈ 10 × 10 -6 / K) is added between the substrate and the HEA plastic layer to alleviate the CTE mismatch.

[0120] Step 3.5: Calculate the phase stability between the substrate and the biomimetic gradient coating under high temperature environment.

[0121] Step 3.5 calculates the phase stability to finally verify the feasibility of the coating optimization scheme, ensuring that the optimization scheme is stable in the actual application environment, and that the biomimetic gradient coating does not precipitate brittle phases (such as Laves phase, sigma phase) above 1000°C. The final optimization design scheme is output.

[0122] Specifically, Thermo-Calc is used to calculate the phase stability of the substrate-HEA plastic layer, HEA / HEC toughening layer, and nanocrystalline HEC superhard layer at different temperatures.

[0123] G 涂层 (T) = H - TS (6)

[0124] In the formula: H represents enthalpy; TS represents the entropy term, which ensures that the coating does not produce non-equilibrium phases.

[0125] Calculate whether sigma phase precipitates above 1000°C:

[0126] ΔG σ = G HEA - G σ (7)

[0127] If ΔG σ > 0, the sigma phase may precipitate, and the composition needs to be optimized.

[0128] Optimization composition, for example: reduce Cr content to avoid sigma phase precipitation, control Cr content at 15%-20% to reduce sigma phase formation. Or optimize Nb / Ti content to improve high temperature stability, increase Nb content (>10%) to stabilize high-entropy alloy structure. Or optimize the high-temperature phase transition of the gradient coating, calculate the Gibbs free energy under different Ti / Nb combinations to ensure that the coating has no brittle phase precipitation at 1200°C.

[0129] Through the optimization process of steps 3.2-3.5, the biomimetic gradient coating can have superior performance such as high bonding force, low thermal stress, strong thermal stability, and high temperature resistance.

[0130] Step 4: Build an atomic model reflecting the interface structure of the high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic composite toughening layer, and nanocrystalline high-entropy ceramic superhard layer. Under the set temperature, pressure, and boundary conditions, perform molecular dynamics simulation on the atomic system to obtain data on lattice and grain parameters, grain boundary width, and phase proportion changes with position. Based on the obtained data, determine a gradual transition scheme for the microstructure to achieve smooth connection between the layers of the coating in terms of lattice, grain, grain boundary, and phase structure.

[0131] The purpose of step 4 is to build an atomic scale model through molecular dynamics simulation to reveal the crystal lattice structure, grain boundary distribution and local changes in phase composition at the interface of each layer. The microscopic transition characteristics at the interface are clarified to provide a basis for designing a microstructure scheme with a smooth transition, thereby improving the bonding strength and crack resistance of the interface. This step lays the foundation for the continuous and smooth structural transition of the coating at the atomic level, reducing the weak bonding problem caused by microscopic discontinuity.

[0132] Specifically, step 4 includes:

[0133] Step 4.1: Building an atomic model. Step 4.1 includes:

[0134] Step 4.1.1: Material selection and crystal structure determination. According to the specific materials of the high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic composite toughening layer, and nanocrystalline high-entropy ceramic superhard layer (such as CoCrFeNiTi and CoCrFeNi-TiC / NbC), the respective lattice parameters, crystal structures and defect information are obtained.

[0135] Step 4.1.2: Interface region model design. In the modeling process, atomic arrangement models between adjacent two layers (such as high-entropy alloy plastic layer and composite toughening layer, composite toughening layer and superhard layer) are constructed, focusing on the interface atomic arrangement, interface transition zone width, and possible grain boundaries and phase mixing regions.

[0136] Referring to the drawings accompanying the specification Figure 4 , Figure 4 The distribution, size and geometry of the grains 9 and grain boundaries 10 in the microstructure layer are shown. In addition, the specific arrangement of the crystal lattice and the microstructure can also be analyzed in high-resolution microscopy or other nanoscale analysis methods. Through molecular dynamics simulation, the influence of these microstructure factors on the interface damage behavior such as lattice distortion, stress concentration and cracks within the coating and between interfaces can be analyzed.

[0137] Step 4.1.3: Multi-scale model integration. For local regions, a representative atomic model is constructed using periodic boundary conditions, while for interface regions, a larger non-periodic model is constructed to fully capture local structural discontinuities and lattice distortions.

[0138] Step 4.2: Setting molecular dynamics simulation parameters. Step 4.2 includes:

[0139] Step 4.2.1: Selection of potential function. According to different material systems, suitable interatomic potential (such as EAM potential, MEAM potential or Tersoff potential) is selected.

[0140] Step 4.2.2: Initial conditions and temperature control. Set initial atomic positions and velocities, use energy minimization method to obtain the equilibrium configuration of the system, and use the Nosé-Hoover temperature controller to realize constant temperature or temperature gradient conditions in the simulation, simulate the actual preparation or heat treatment process.

[0141] Step 4.2.3: Simulation steps and time scales. Refine the time step and determine the total number of steps to capture the interface atomic rearrangement, grain boundary diffusion and local phase transition process, and use multi-step simulation (such as energy minimization first, then temperature rise, holding, and cooling simulation) if necessary to reflect the real process.

[0142] Step 4.3: Data collection and analysis. Step 4.3 includes:

[0143] Step 4.3.1: Analysis of atomic arrangement and local structure. Use radial distribution function (RDF), coordination number statistics, atomic displacement vector analysis and other methods to quantitatively describe the continuity of atomic arrangement and local defect distribution in the interface region.

[0144] Step 4.3.2: Calculate interface energy and bonding strength. By calculating the total energy of different interface configurations, compare the interface energy of each scheme, and use the interatomic interaction energy to calculate the interface bonding strength, introduce the interface diffusion coefficient and atomic rearrangement rate as auxiliary indicators.

[0145] Step 4.3.3: Evaluate lattice matching degree and grain boundary width. Use local strain analysis and lattice mismatch parameter calculation to quantitatively evaluate the lattice matching degree at the interface, and further judge the rationality of the transition scheme by statistical grain boundary width and discontinuous area proportion.

[0146] Step 4.4: According to the data analysis results of step 4.3, compare the microstructure of different interface designs (such as transition zone width, element gradient distribution) with multiple schemes, select the design scheme with the highest interface bonding energy and the highest lattice matching degree, and feed back the selected microstructure transition scheme to the overall coating design, adjust the atomic model parameters, and carry out iterative simulation, output the gradual transition scheme of the microstructure.

[0147] Step 5: According to the actual process parameters and the temperature-dependent mechanical properties of each layer of material, establish a finite element model containing the tool substrate and each layer of the gradient coating, apply thermal-mechanical coupling load conditions in the model, simulate the thermal cycle and mechanical load in the working process, obtain the stress, strain and residual stress distribution at the interface, analyze the stress concentration and deformation in the interface region, and determine an adaptive and optimized interface mechanical performance transition design scheme.

[0148] The purpose of Step 5 is to evaluate the mechanical response of the coating interface at the macro scale, especially the stress, strain and residual stress distribution under thermal-mechanical coupling, using finite element simulation. By simulating the mechanical behavior under actual working conditions, stress concentration or weak areas that may occur at the interface are identified, and the interface design is optimized to ensure good mechanical stability of the interface under load. This step provides quantitative analysis and verification of the mechanical properties of the coating design, ensuring the reliability and durability of the overall structure during long-term use.

[0149] Specifically, Step 5 includes:

[0150] Step 5.1: Establish a three-dimensional geometric model containing the tool substrate and each layer of the gradient coating, specifically divide the interface area to form independent interface elements or use bonded contact to clearly define the transition area between different layers.

[0151] Each layer of the gradient coating includes a high-entropy alloy plastic layer, a high-entropy alloy / high-entropy ceramic composite toughening layer, and a nanocrystalline high-entropy ceramic superhard layer.

[0152] Step 5.2: Assign temperature-dependent elastic modulus, Poisson's ratio, thermal expansion coefficient, yield strength and other plasticity parameters to each layer.

[0153] Step 5.3: Reasonably mesh the entire model, use local encryption meshing in the interface and areas with high stress gradients, and determine the optimal mesh size through mesh independence verification.

[0154] Areas with high stress gradients, such as the junction between the coating and the substrate.

[0155] Step 5.4: Apply thermal load according to the actual process flow, simulate the thermal cycle process, and form a temperature gradient.

[0156] Simulating the thermal cycle process includes heating, holding, and cooling.

[0157] Step 5.5: Apply fixed boundary conditions or appropriate constraints to the tool substrate, apply external force or displacement load in the working state to the coating and interface area, simulate the mechanical response generated during machining or use, and set the interface contact properties.

[0158] Interface contact properties such as slip, friction or adhesive failure phenomena.

[0159] Step 5.6: Use a nonlinear thermal-mechanical coupling solver to jointly solve the temperature field and stress field, and ensure convergence through step-by-step solving.

[0160] This step ensures that the material nonlinear behavior and large deformation effects are considered. Step-by-step solving, such as thermal analysis followed by mechanical analysis, or coupled solving.

[0161] Step 5.7: Extract stress, strain, residual stress and local deformation data in the key area, and use post-processing tools to generate temperature field, stress field distribution map and local magnified map to identify the location and magnitude of stress concentration.

[0162] The key areas are mainly the interface areas between layers.

[0163] Step 5.8: Compare the maximum stress, strain distribution and residual stress level of the interface region under different design schemes, establish stress concentration index, and use the construction of objective function (e.g., minimizing the maximum stress at the interface, reducing the stress gradient, and reducing residual stress) as the basis for optimization, and analyze the influence of different interface contact parameters, layer thickness and material transition zone design on the objective function.

[0164] Step 5.9: Perform sensitivity analysis on key parameters to determine their influence on mechanical response, propose design improvement schemes based on the sensitivity analysis results, and optimize the interface transition design.

[0165] Key parameters include interface contact stiffness, interlayer coating thickness, and temperature loading amplitude.

[0166] Step 5.10: Perform multi-condition simulation verification on the optimized design scheme, feed the verification results back into the model, and perform multiple iterative optimizations until an interface mechanical performance transition scheme that meets the requirements of actual working conditions is obtained.

[0167] The purpose of multi-condition simulation verification is to ensure that the interface can maintain low stress concentration and stable mechanical properties under different thermal and mechanical loads.

[0168] In step 5.10, if there is experimental data to support the comparison, the simulation results can be compared with the experimental data, and the model parameters can be corrected if necessary.

[0169] Step 6: Preprocess and extract features from the obtained material composition, microstructure and mechanical properties data of each layer to form a multidimensional dataset. Use BP neural network and particle swarm optimization algorithm to train and optimize the data, construct a mathematical model that reflects the interrelationship of composition, structure and properties between the layers of the gradient coating, and embed the mathematical model into the MATLAB mathematical modeling platform to realize the gradual transition design between the tool substrate and the gradient coating and dynamically control the interface parameters.

[0170] The purpose of step 6 is to integrate the aforementioned theoretical calculations and simulation data using machine learning techniques such as BP neural networks and particle swarm optimization to establish a three-in-one synergistic control model of "composition entropy - structural entropy - performance entropy" for gradient coatings. This model reflects the intrinsic relationship between material composition, microstructure, and mechanical properties. It enables precise design of the gradual transition between the tool substrate and the gradient coating, dynamically controls interface parameters, and optimizes the overall interfacial bonding strength and toughness. This model, through data-driven optimization methods, provides theoretical support and optimization strategies for the multi-objective synergistic design of coatings, significantly improving overall coating performance and providing reference and improvement directions for the design of similar systems in the future.

[0171] Specifically, step 6 includes:

[0172] Step 6.1: Collect the multi-scale data obtained in Steps 3-5, remove missing values, handle outliers and normalize the original data, use principal component analysis (PCA) or feature extraction techniques to screen key indicators related to interface bonding strength, toughness and overall performance, and construct the input data matrix as the basis for subsequent models.

[0173] Multi-scale data includes thermo-mechanical coupling simulation data, molecular dynamics simulation data, and finite element simulation data. Specifically, it includes, but is not limited to: material composition, layer thickness, and interface element distribution data; lattice parameters, grain boundary width, local defect information, and atomic rearrangement data of the interface region; and mechanical response data such as stress, strain, residual stress distribution, and stress concentration indices at the interface.

[0174] Data cleaning and normalization of the raw data can ensure that all variables are on the same order of magnitude, which facilitates subsequent feature extraction and model training.

[0175] Step 6.2: Define component entropy, structural entropy, and performance entropy, and construct a unified index system based on the entropy values ​​of component entropy, structural entropy, and performance entropy. Establish a multi-objective optimization model and set the collaborative regulation objective as follows:

[0176] minJ=ω1·f(S c )+ω2·f(S s )+ω2·f(S p )

[0177] In the formula: ω1, ω2, and ω2 are the weighting factors for component entropy, structural entropy, and performance entropy, respectively; f(S c f(S) represents the cost function for the deviation of various indices of component entropy from the ideal state; s f(S) represents the cost function for the deviation of various indices of structural entropy from the ideal state; p ) represents the cost function for the deviation of various performance entropy indicators from the ideal state.

[0178] Compositional entropy reflects the uniformity and diversity of the composition and elemental distribution of each layer of material; its calculation formula can be found in the information entropy model. Structural entropy describes the degree of order in lattice matching, grain boundary width, and defect distribution within the interface microstructure. Performance entropy quantifies the balance and robustness of interface mechanical properties (such as interfacial bond strength, residual stress, and strain distribution).

[0179] The purpose of constructing a unified index system from the above entropy values ​​is to characterize the overall synergistic control effect of the gradient coating.

[0180] Step 6.3: Select a multi-layer feedforward neural network as the basic model structure. The input layer receives the preprocessed component, structure and performance data, and the output layer outputs the strength and toughness prediction values. The number of hidden layers and nodes is set according to the data complexity, and activation functions (such as ReLU or sigmoid) are used to improve the model's nonlinear mapping ability.

[0181] Step 6.4: Train the network using the backpropagation algorithm (BP algorithm), set the loss function to mean squared error or other suitable indicators, evaluate the model's generalization ability using cross-validation, adjust the network parameters until both the training error and validation error meet the expected requirements, implement the training process using the MATLAB mathematical modeling platform, and generate a training report and performance evaluation charts.

[0182] Step 6.5: Perform a global search of the weights, biases, and hyperparameters of the BP neural network using the Particle Swarm Optimization (PSO) algorithm to further reduce training error.

[0183] The process of globally searching the weights, biases, and hyperparameters of a BP neural network using the Particle Swarm Optimization (PSO) algorithm includes: initializing a certain number of particles (each particle represents a set of network parameters), defining the fitness function as the network's prediction error on the validation set, updating the velocity and position of each particle, iterating until convergence, finally obtaining the optimal parameter set, and applying it to the BP neural network model to form a comprehensively optimized collaborative control model.

[0184] Step 6.6: Integrate the “composition entropy-structure entropy-performance entropy” collaborative regulation model, which has undergone data preprocessing, BP neural network training, and PSO tuning, into the MATLAB platform to establish a systematic simulation calculation framework. Validate the model using historical data and newly generated simulation data to ensure that the model prediction results match the actual performance indicators. Analyze the prediction error and sensitivity, and adjust feature engineering, network structure, and optimization parameters to form a closed-loop improvement process.

[0185] Once a systematic simulation framework is established, it can be used to predict the synergistic performance of different coating design schemes and guide the adjustment of actual process parameters. Based on real-time data obtained during actual processing and use, the training dataset is continuously updated, and the model is dynamically corrected using an online learning mechanism to ensure long-term stability and adaptability.

[0186] The above descriptions are merely embodiments of the present invention, and common knowledge regarding specific structures and characteristics of the solutions is not described in detail here. It will be apparent to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the present invention is defined by the appended claims rather than the foregoing description. Therefore, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A cross-scale biomimetic gradient coating design method based on multi-principal high-entropy systems, characterized in that, The method includes: Step 1: Sequentially form a high-entropy alloy plastic layer, a high-entropy alloy / high-entropy ceramic toughening layer, and a nanocrystalline high-entropy ceramic superhard layer on the surface of the tool substrate to construct a biomimetic gradient coating with a gradient transition structure; Step 2: Determine the material composition of the tool substrate and each layer of the gradient coating, so that the material of the high-entropy alloy plastic layer in the gradient coating matches the high-entropy alloy bonding in the tool substrate. Select the material of the nanocrystalline high-entropy ceramic superhard layer according to the target machining object of the tool. Based on the material of the high-entropy alloy plastic layer and the material of the nanocrystalline high-entropy ceramic superhard layer, determine the material of the high-entropy alloy / high-entropy ceramic composite toughening layer. Step 3: The interfacial properties of the biomimetic gradient coating are analyzed using a thermo-mechanical coupling calculation method. Based on the calculation and analysis results, the gradient transition design of the interfacial components is optimized, and the gradual transition scheme of the interfacial material components is determined. Step 4: Construct an atomic model reflecting the interface structure of the high-entropy alloy plastic layer, the high-entropy alloy / high-entropy ceramic composite tough layer, and the nanocrystalline high-entropy ceramic superhard layer. Under set temperature, pressure, and boundary conditions, perform molecular dynamics simulation on the atomic system to obtain data on the changes of lattice and grain parameters, grain boundary width, and phase ratio with position. Based on the obtained data, determine a gradual transition scheme for the microstructure so that the coating layers can achieve smooth connection in terms of lattice, grain, grain boundary, and phase structure. Step 5: Based on the actual process parameters and the temperature-dependent mechanical properties of each layer of material, establish a finite element model including the tool substrate and each layer of the gradient coating. Apply thermo-mechanical coupled load conditions to the model to simulate the thermal cycle and mechanical load during the working process, obtain the distribution of stress, strain and residual stress at the interface, analyze the stress concentration and deformation in the interface area, and determine an adaptive and optimized interface mechanical property transition design scheme. Step 6: Preprocess and extract features from the obtained material composition, microstructure and mechanical properties data of each layer to form a multidimensional dataset. Use BP neural network and particle swarm optimization algorithm to train and optimize the data, construct a mathematical model that reflects the interrelationship of composition, structure and properties between the layers of the gradient coating, and embed the mathematical model into the MATLAB mathematical modeling platform to realize the gradual transition design between the tool substrate and the gradient coating and dynamically control the interface parameters.

2. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, In step 2, the material of the tool substrate includes, but is not limited to, any one of TiC-CrFeCoNi, Ti(C,N)-CoCrCuFeNi, and TiC-CrMnFeCoNi.

3. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, In step 2, the materials of the high-entropy alloy plastic layer include, but are not limited to, any one of CrFeCoNi, CoCrCuFeNi, and CrMnFeCoNi; the materials of the nanocrystalline high-entropy ceramic superhard layer include, but are not limited to, any one of AlTiSiN, (TiZrNbTaMo)C / N, and (TiAlTaCrZr)N; and the materials of the high-entropy alloy / high-entropy ceramic composite toughening layer include, but are not limited to, any one of (CrFeCoNi)-AlTiSiN and (CoCrCuFeNi)-(TiZrNbTaMo)C.

4. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, Step 3 includes: Step 3.1: Select a database suitable for biomimetic gradient coatings in Thermo-Calc software, establish a biomimetic gradient coating structure model, and set the required simulation environment parameters; Step 3.2: Calculate the interfacial bonding energy of the matrix, high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic toughening layer, and nanocrystalline high-end ceramic superhard layer, and adjust the composition of the interfacial materials according to the calculation results; Step 3.3: Under thermal cycling and mechanical load conditions, calculate the interfacial residual stress of each layer of the biomimetic gradient coating; Step 3.4: Calculate the difference in thermal expansion coefficients between the substrate and the high-entropy alloy plastic layer in the biomimetic gradient coating, as well as between adjacent layers in the biomimetic gradient coating, and optimize the interface thermal matching accordingly; Step 3.5: Calculate the phase stability between the substrate and the biomimetic gradient coating under high temperature conditions.

5. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, In step 2, the material of the tool substrate includes a composite high-entropy alloy binder phase and a ceramic hard phase, wherein the high-entropy alloy binder phase in the tool substrate material is the same as the high-entropy alloy binder phase in the high-entropy alloy plastic layer material.

6. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, Step 5 includes: Step 5.1: Establish a three-dimensional geometric model including the tool substrate and each layer of the gradient coating, divide the interface area into independent interface units or use adhesive contact, and clarify the transition area between different layers. Step 5.2: Assign temperature-dependent elastic modulus, Poisson's ratio, coefficient of thermal expansion, yield strength and other plastic parameters to each layer; Step 5.3: Perform reasonable meshing on the entire model, and use local mesh refinement in areas with high stress gradients and interfaces. Determine the optimal mesh size through mesh independence verification. Step 5.4: Apply thermal load according to the actual process flow to simulate the thermal cycle process and form a temperature gradient; Step 5.5: Apply fixed boundary conditions or appropriate constraints to the tool substrate, apply external forces or displacement loads under working conditions to the coating and interface areas, simulate the mechanical response generated during machining or use, and set the interface contact properties at the same time. Step 5.6: Use a nonlinear thermo-mechanical coupled solver to solve the temperature field and stress field together, and solve in steps to ensure convergence; Step 5.7: Extract stress, strain, residual stress, and local deformation data in key areas, and use post-processing tools to generate temperature field, stress field distribution maps, and local magnified maps to identify stress concentration locations and amplitudes; Step 5.8: Compare the maximum stress, strain distribution and residual stress level of the interface area under different design schemes, establish stress concentration index, and analyze the influence of different interface contact parameters, layer thickness and material transition zone design on the objective function by constructing an objective function as the basis for optimization. Step 5.9: Conduct sensitivity analysis on key parameters to determine their influence on mechanical response, propose design improvement schemes based on the sensitivity analysis results, and optimize interface transition design; Step 5.10: Perform multi-condition simulation verification on the optimized design scheme, feed the verification results back into the model, and perform multiple iterative optimizations until an interface mechanical performance transition scheme that meets the requirements of actual working conditions is obtained.

7. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, Step 6 includes: Step 6.1: Collect the multi-scale data obtained in Steps 3-5, remove missing values, handle outliers and normalize the original data, use principal component analysis or feature extraction techniques to screen key indicators related to interface bonding strength, toughness and overall performance, and construct the input data matrix as the basic data for subsequent models. Step 6.2: Define component entropy, structural entropy, and performance entropy, and construct a unified index system based on the entropy values ​​of component entropy, structural entropy, and performance entropy. Then, build a multi-objective optimization model and set collaborative control objectives. Step 6.3: Select a multi-layer feedforward neural network as the basic model structure. The input layer receives the preprocessed component, structure and performance data, and the output layer outputs the strength and toughness prediction values. The number of hidden layers and nodes is set according to the data complexity, and activation functions are used to improve the model's nonlinear mapping capability. Step 6.4: Train the network using the backpropagation algorithm, set the loss function to mean squared error or other suitable indicators, evaluate the model's generalization ability using cross-validation, adjust the network parameters until both the training error and validation error meet the expected requirements, implement the training process using the MATLAB mathematical modeling platform, and generate a training report and performance evaluation charts. Step 6.5: Use the PSO algorithm to perform a global search on the weights, biases, and hyperparameters of the BP neural network to further reduce training error; Step 6.6: Integrate the "composition entropy-structure entropy-performance entropy" collaborative regulation model, which has undergone data preprocessing, BP neural network training, and PSO tuning, into the MATLAB platform to establish a systematic simulation calculation framework. Validate the model using historical data and newly generated simulation data to ensure that the model prediction results match the actual performance indicators. Analyze the prediction error and sensitivity, and adjust feature engineering, network structure, and optimization parameters to form a closed-loop improvement process.

8. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 1, characterized in that, Step 4, the process of generating the optimal microstructure gradient transition scheme includes: First, based on the crystal structure, lattice parameters, and defect information of the selected high-entropy alloy plastic layer, high-entropy alloy / high-entropy ceramic composite tough layer, and nanocrystalline high-entropy ceramic superhard layer, an atomic model that meets the requirements of periodic and non-periodic boundary conditions is constructed, with a focus on designing the atomic arrangement and transition zone of the interface region; then, based on the characteristics of each material system, an appropriate interatomic interaction potential is selected, initial atomic positions and velocities are set, and the system equilibrium configuration is obtained using the energy minimization method; then, a Nosé-Hoover temperature controller is used to perform isothermal or temperature gradient control under simulated real heat treatment conditions, and the interface atomic rearrangement, grain boundary diffusion, and local phase transition phenomena are captured by refining the time step and multi-step simulation; finally, the simulation data is collected and quantitatively evaluated using methods such as radial distribution function, coordination number statistics, and local strain analysis, and the interface energy, bonding strength, and lattice matching degree are calculated to determine the optimal microstructure gradient transition scheme.

9. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 7, characterized in that, In step 6.5, the process of globally searching the weights, biases, and hyperparameters of the BP neural network using the PSO algorithm includes: initializing a certain number of particles, defining the fitness function as the prediction error of the network on the validation set, updating the velocity and position of each particle, iterating the search until convergence, finally obtaining the optimal parameter set, and applying it to the BP neural network model to form a comprehensively optimized collaborative control model.

10. The cross-scale biomimetic gradient coating design method based on a multi-principal high-entropy system according to claim 7, characterized in that, In step 6.2, composition entropy is used to reflect the uniformity and diversity of the composition and element distribution of each layer of material, and the calculation formula can refer to the information entropy model; structural entropy is used to describe the degree of order of lattice matching, grain boundary width and defect distribution in the microstructure of the interface; performance entropy is used to quantify the balance and robustness of the mechanical properties of the interface.

Images