A membrane rib composite structure bias filter target plate

Abstract

The application belongs to the technical field of key components of magnetic confinement nuclear fusion reactors, and discloses a membrane-rib composite structure divertor target plate, which comprises a plasma-facing material layer, a membrane-rib composite structure heat sink layer and a backplate support layer. The core of the application is the membrane-rib composite structure heat sink layer, which comprises a rib structure for disturbing fluid to promote boiling heat exchange, preferably, the rib structure adopts a bionic veinlet structure design; and a porous functional membrane for realizing efficient separation of gas and liquid, preferably, the functional membrane is a SiO2 porous membrane. In addition, an embedded optical fiber sensor is integrated in the backplate support layer of the target plate to form an intelligent diagnosis system for real-time monitoring of the temperature and strain of the target plate. Through the innovative bionic structure design and structure-function integrated design, the heat exchange capacity and operation reliability of the divertor target plate are significantly improved, which provides key technical support for the stable operation of future fusion reactors.

Description

Technical Field

[0001] This invention relates to the technical field of key components for magnetic confinement nuclear fusion reactors, and more specifically, to a high-efficiency heat exchange component for future fusion reactors, particularly a divertor target plate with a membrane rib composite structure. Background Technology

[0002] Nuclear fusion energy, as a clean energy source with enormous potential, has entered a new stage of development. The divertor target plate is one of the most critical core components in the International Thermonuclear Experimental Reactor (ITER) and future commercial fusion reactors. It directly faces and withstands extreme heat loads and strong particle stream impacts of up to 10-20 MW / m² from the core plasma. Therefore, the thermal load-bearing capacity, structural stability, and manufacturing level of the divertor target plate directly determine the operational stability and reliability of the entire fusion device.

[0003] Currently, divertor target plate technology still faces severe challenges. First, under extreme heat loads, traditional active cooling structures (such as straight channels and U-shaped water-cooled channels) are approaching their physical limits, exhibiting bottlenecks such as uneven flow fields, localized heat transfer dead zones, and large pressure drops, limiting the potential for improving heat transfer performance. Second, fusion reactors require components to have service lives of up to 100,000 hours, but under the combined effects of high temperatures, strong radiation, and thermal cycling, the performance of target plate materials degrades significantly, and there is a lack of effective in-situ, real-time health monitoring methods, leading to a large deviation between predicted divertor performance and actual service life, resulting in high safety risks.

[0004] Therefore, developing a novel divertor target plate structure with a completely new enhanced heat exchange mechanism and integrated intelligent diagnostic function is of great significance for promoting the commercial application of nuclear fusion energy. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a divertor target plate with a membrane rib composite structure to solve the technical problems of the heat exchange efficiency bottleneck of existing target plates under extreme heat loads and the lack of effective service condition monitoring methods.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A membrane rib composite structure divertor target plate, characterized in that the target plate is a multi-layer composite structure, comprising, from the surface bearing the plasma heat flow inward: a plasma-facing material layer, a membrane rib composite structure heat sink layer, and a backplate support layer; wherein the plasma-facing material layer, the membrane rib composite structure heat sink layer, and the backplate support layer are integrally connected by a stacking forging process.

[0007] The plasma-facing material layer is a flat or tile-like structure made of tungsten or tungsten alloy material. The surface facing the plasma can be configured with macroscopic or microscopic textures as needed to regulate the interaction between the plasma and the material.

[0008] The backplate support layer is a plate-shaped or shell-shaped structure made of stainless steel or high-temperature alloy material, with a groove pre-set inside to accommodate the fiber optic sensor, providing structural support and rigidity for the entire target plate.

[0009] The heat sink layer of the membrane rib composite structure is made of copper or copper-chromium-zirconium alloy and has a cooling channel inside. The cooling channel integrates a biomimetic leaf vein rib structure and a vapor-liquid separation functional membrane, which together constitute a membrane rib composite cooling structure for enhancing heat exchange.

[0010] The biomimetic leaf vein rib structure is a flow-disrupting structure set on the inner wall of the cooling channel. Its geometry is inspired by the natural leaf vein network, with a lateral parallel vein structure. That is, the lateral veins branching from the main vein are parallel to each other, and the lateral veins form a certain angle with the main vein. This structure aims to disturb the fluid in the near-wall region, disrupt the laminar boundary layer, increase the vaporization nuclei, and promote efficient subcooled boiling.

[0011] The vapor-liquid separation functional membrane is a porous SiO2 membrane disposed on the heated wall surface within the cooling channel. This membrane is fixed to the heated wall surface through a stacking and forging process. Its function is to continuously wet the heated wall surface with the liquid coolant through capillary effect, while simultaneously blocking bubbles generated during boiling and guiding them to the center of the channel to be carried away by the main flow, thereby preventing the occurrence of the critical heat flux (CHF) phenomenon.

[0012] The target plate also includes an integrated intelligent diagnostic system, which comprises a fiber optic sensor network embedded in the groove of the backplate support layer, a fiber optic demodulator for signal analysis, and a host computer for data processing and health assessment. The fiber optic sensors are embedded in the backplate support layer using an ultrasonic consolidation process.

[0013] Meanwhile, this invention also discloses a design method for a membrane rib composite structure divertor target plate. This method adopts biomimetic principles and multi-scale optimization technology, and its process includes: 1) Performance modeling and analysis: First, a geometric model of the membrane-rib composite structure is established. By observing parallel leaf vein structures in nature, key leaf vein structural features are extracted, such as lateral vein height (…). ), lateral vein width ( ), lateral vein spacing ( The key geometric features of the leaf veins include: the angle (θ) between the lateral veins and the main vein; the angle between the main vein and the lateral vein; the angle between adjacent lateral veins ...

[0014] Secondly, an equivalent analysis method for the macroscopic performance of periodic membrane rib units is established using the asymptotic homogenization theory. The specific implementation steps are as follows: a) Extract a representative, periodically repeatable minimum structural unit from the geometric model as the representative volume element (RVE). b) Apply periodic boundary conditions (such as unit temperature difference and unit strain) to the RVE to simulate its stress and heat transfer environment in the macroscopic continuum; c) Solve the heat flux density and stress-strain field response of the RVE under specific boundary conditions (such as unit temperature difference and unit strain) using finite element analysis (FEA); d) Based on the microscopic field quantities obtained from the solution, calculate the macroscopic equivalent performance parameters using the volume averaging method. For example, the equivalent thermal conductivity ( ), equivalent elastic modulus ( ) and equivalent thermal expansion coefficient ( The calculation of ε is based on assigning the properties of representative microscopic units back to the macroscopic level as average equivalent values ​​of the macroscopic level. Specifically, for any macroscopic physical quantity E and its corresponding microscopic quantity ε, the relationship is as follows:

[0015] in, The calculation represents the volume average within the RVE region. This step equates the membrane-ribbed composite heat sink layer with its complex microstructure to a homogeneous material layer with anisotropic macroscopic properties, providing input parameters for subsequent thermo-mechanical analysis of the entire target plate.

[0016] Furthermore, to accurately analyze the stress concentration and potential cracking problems at the interfaces of multi-layered heterogeneous materials (such as W / CuCrZr, CuCrZr / SS316L) in the target plate, the interpolation functions for the displacement and temperature fields at the multi-material interfaces of the target plate element are established based on the extended finite element method (XFEM):

[0017] in, Let a point be a point within the computational domain; These are standard finite element shape functions; For standard node degrees of freedom, This represents the standard node set; the first term is the standard finite element displacement approximation. The second and third terms are enrichment terms. It is the Heaviside function that describes the displacement discontinuity on both sides of the crack surface. These are the degrees of freedom of the corresponding enriched nodes. It is the set of nodes it influences; It is a family of analytic functions describing the singularity of stress at the crack tip. The term number of this analytic function family, taking values ​​from 1 to 4. These are the degrees of freedom of the corresponding enriched nodes. It is the set of nodes that it affects. This method allows for the accurate simulation of the initiation and propagation of interface cracks without relying on mesh generation.

[0018] A cohesive constitutive model (CZM) is used to describe the mechanical properties at the material interfaces of a component. This model characterizes the damage evolution process at the interface under normal and tangential stresses by defining a traction-separation constitutive relation (TSL). This relation is typically bilinear or exponential in form, and its key parameters include the interfacial tensile strength (TST). ), shear strength ( ) and fracture energy ( ).

[0019] Through the above modeling and analysis, the inherent mapping law between material parameters, structural form, and comprehensive performance has been clarified, specifically including: a) Coefficient of thermal expansion mismatch ( ) and peak interfacial shear stress ( The correlation is approximately linear and positive, which can be expressed as: ,in, For geometry-related constants, For the equivalent elastic modulus, For working temperature difference; b) Rib height ,width and spacing With equivalent heat transfer coefficient and flow pressure drop A multi-objective optimization surface is constructed, and its relationships are obtained through a series of parametric simulation calculations and fitted as a response surface function. ; c) Interfacial bonding strength (from the cohesive force model) and The characterization directly determines the fatigue life of the target plate under thermal cycling. Through an extended form of Paris's law, the range of the interface stress intensity factor ( ) and crack propagation rate ( This correlation enables quantitative prediction of thermal cycling fatigue life.

[0020] 2) Multi-scale optimized biomimetic design: Based on the mapping rules clearly defined in step 1) above, i.e., structural parameters (such as...) , , ) and performance indicators (such as , , The quantitative relationship between these factors is established to create a multi-objective optimization design model encompassing performance, function, and manufacturing constraints. This model aims to maximize heat transfer efficiency and minimize structural stress, while using material usage, fluid pressure drop, and the minimum feature size and maximum overhang angle in additive manufacturing processes as constraints. This multi-objective optimization model can be expressed as:

[0021] in, Let x be the objective function, and let x be the design variable (such as the height, width, and spacing of the ribs). For heat exchange efficiency, For maximum stress, for constraint functions and for .

[0022] The non-dominated sorting genetic algorithm (NSGA-II) is used to solve this multi-objective optimization model. The specific solution steps are as follows: a) Initialization: Randomly generate an initial population containing N individuals (i.e., N sets of design variables x); b) Iterative optimization: Perform non-dominated sorting and crowding calculation on the current population, and then generate offspring populations through selection, crossover, and mutation operations; c) Sensitivity analysis assistance: To improve convergence efficiency, sensitivity analysis techniques based on the adjoint variable method can be introduced to calculate the gradients of the objective function and constraint function with respect to the design variables. This gradient information can be used to guide crossover and mutation operations, making them more directional in exploring the optimization direction; d) Merging and screening: Merge the parent and offspring populations, perform non-dominated sorting and crowding calculation again, and select a new set of N optimal individuals to form the next generation population; e) Termination: Repeat steps b) to d) until the preset number of iterations is reached or the population converges. Finally, a set of Pareto optimal solutions is obtained, which represents a series of optimal solutions that balance heat transfer efficiency and structural reliability. Those skilled in the art can select an optimal biomimetic configuration of the membrane rib composite structure from this solution set according to specific engineering requirements. For example, a configuration that maximizes heat transfer efficiency while ensuring that the stress is below the allowable value. That is, the final preferred configuration parameters as described in the following embodiments are: rib height 6mm, rib thickness 3mm, and rib spacing 3mm.

[0023] Compared with the prior art, the present invention has the following significant advantages: 1. Extremely high heat transfer performance: A novel vapor-liquid separation and boiling heat transfer mechanism is achieved through a membrane rib composite structure, effectively avoiding the CHF phenomenon and breaking through the physical limits of traditional enhanced heat transfer methods. It is expected to withstand a high heat load of no less than 20MW / m² under low pressure (≤6 atm) and low flow rate (≤5L / min) water cooling conditions, while keeping the maximum surface temperature of the target plate below 1500℃.

[0024] 2. Integrated structure and function: The fiber optic sensor network is embedded into the target plate structure with low damage through ultrasonic consolidation process, realizing a high degree of integration between heat exchange function and structural health monitoring function. This makes it possible to obtain the real-time service status of the divertor, which greatly improves the safety, stability and reliability of the fusion device operation.

[0025] 3. Advanced Design and Manufacturing Methods: A design method based on bionics and multi-scale optimization was proposed, achieving systematic optimization from materials and structure to performance. Combined with advanced manufacturing technologies such as stacked forging and ultrasonic consolidation, a feasible technical path was provided for the mass production and low-cost manufacturing of high-performance divertor targets. Attached Figure Description

[0026] To more clearly illustrate the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0027] Figure 1 This is a schematic diagram of the design process of the membrane rib composite structure based on bionic principles in this invention.

[0028] Figure 2 This is a schematic diagram of the integrated method for the target plate intelligent diagnostic system based on embedded optical fiber in this invention.

[0029] Figure 3 This is a schematic diagram of the process for low-damage embedding of the fiber optic sensor into the backplane support layer in this invention.

[0030] Figure 4 This is a schematic diagram showing the overall structure and exploded view of the membrane rib composite structure divertor target plate unit proposed in this invention.

[0031] The labels in the diagram are explained as follows: 1. Plasma-facing material layer; 2. Membrane-ribbed composite heat sink layer; 3. Backplate support layer; 4. Bionic leaf vein rib structure; 5. Gas-liquid separation SiO2 membrane; 6. Fiber optic sensor. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0033] This embodiment provides a complete preparation and performance testing process for a divertor target plate with a membrane rib composite structure, aiming to verify the feasibility and superiority of the technical solution of the present invention.

[0034] 1. Target plate structure design First, according to the design method described in the invention, a divertor target plate with a heating surface size of 20mm × 20mm is designed, such as... Figure 4 The specific design parameters are as follows: Plasma-facing material layer (1): Pure tungsten (W) material with a thickness of 2 mm.

[0035] The heat sink layer of the membrane rib composite structure (2) is made of copper chromium zirconium (CuCrZr) alloy with a thickness of 8mm. The specific parameters of the internal biomimetic leaf vein rib structure are as follows: rib thickness 3mm, rib spacing 3mm, and rib height 6mm.

[0036] Back panel support layer (3): Made of 316L stainless steel with a thickness of 3mm.

[0037] 2. Preparation of key components (1) Preparation of SiO2 membrane (5) for vapor-liquid separation: 1) Preparation of precursor solution: Dissolve polyvinyl alcohol (PVA) in deionized water to prepare a PVA solution with a mass fraction of 8-12%; simultaneously, under the action of an acidic catalyst (such as hydrochloric acid), mix tetraethoxysilane (TEOS), anhydrous ethanol, and water in a molar ratio of 1:4:4 to prepare a silane sol. Then, add the silane sol dropwise to the PVA solution in a mass ratio of 1:1 to 1:3, and mix thoroughly with magnetic stirring to obtain the precursor solution.

[0038] 2) Electrospinning and post-treatment: The precursor solution is loaded into a syringe and spun in an electrospinning apparatus. The process parameters are set as follows: voltage 15-25kV, spinning distance 10-20cm, and solution delivery flow rate 0.5-1.5mL / h. The collected PVA / SiO2 composite fiber felt is dried and then heated to 600-800℃ in a muffle furnace at a heating rate of 2-5℃ / min, and calcined for 2-4 hours to completely remove the PVA organic components, finally obtaining a porous SiO2 membrane with a purity higher than 99.5%.

[0039] 3) Pore size control: By adjusting the ratio of PVA to silane sol in step 1), the spinning process parameters in step 2), and the calcination temperature and time, the average pore size and porosity of the SiO2 film can be precisely controlled, and SiO2 films with an average pore size ≤10µm, a thickness in the range of 50-120µm, and a heat resistance temperature of over 1000℃ can be prepared.

[0040] (2) Fabrication of the backplane support layer (3) for embedded optical fibers: Low-damage embedding process: Ultrasonic bonding technology is used to stack and bond layers of metal foil (such as stainless steel foil) with a thickness of 0.1-0.2 mm. First, grooves matching the fiber diameter are pre-machined into the foil. The fiber optic sensor is placed into the grooves, and then subsequent foil layers are stacked until the entire backplane support layer is manufactured. Figure 3 By optimizing process parameters such as ultrasonic amplitude, consolidation pressure, and welding speed, the metal foil undergoes plastic flow in a low-temperature solid state and fills the area around the optical fiber, achieving low-damage embedding of the optical fiber and ensuring that the loss rate of the embedded sensor is ≤10%.

[0041] 3. Integrated manufacturing of the target plate The three main parts of the target plate were integrally formed using electric field-assisted stacking forging 3D printing technology. The specific process is as follows: First, the prefabricated backplate support layer (3) with embedded optical fiber is used as the substrate.

[0042] On top of this, a membrane rib composite heat sink layer (2) is constructed layer by layer by stacking CuCrZr alloy sheets and performing electric field-assisted stacking forging. Among them, the pre-prepared vapor-liquid separation SiO2 membrane (5) is placed under the heated wall surface and fixed by subsequent stacking.

[0043] Finally, on top of the heat sink layer, a plasma-oriented material layer of tungsten (W) material is constructed using the same layered stacking and electric field-assisted stacking forging process (1).

[0044] The electric field-assisted stacking forging process parameters are: pulse current density 500 A / mm², single pulse duration 100 ms, and pressure 50 MPa. This process utilizes the Joule heating and non-thermal effects of the pulsed electric field to achieve high-strength metallurgical bonding of W / CuCrZr and CuCrZr / SS316L heterojunction interfaces at relatively low macroscopic temperatures. The prepared target plate samples were tested and found to have good interlayer bonding, with a tensile strength coefficient (relative to low-strength materials) reaching 0.85 and a shear strength coefficient reaching 0.92.

[0045] 4. Performance Testing and Verification (1) High heat load test: The prepared target plate sample was mounted on a high-heat-load testing platform (such as an electron beam or ion beam bombardment platform). The cooling water conditions were set as follows: pressure 5 atm, flow rate 4 L / min. A uniform heat load of 20 MW / m² was applied to the heated surface of the target plate. The surface temperature was monitored by an infrared thermal imager, and the internal temperature and strain of the target plate were monitored by an embedded fiber optic sensor (6).

[0046] (2) Establishment and verification of intelligent diagnostic system: 1) Intelligent Diagnostic System Construction: The embedded fiber optic sensor network can collect temperature and strain data at different locations inside the target plate in real time. The data is analyzed by a fiber optic demodulator and transmitted to a host computer. Machine learning algorithms (such as neural networks or support vector machines) are used to train and analyze the massive amounts of collected data to establish a target plate health status assessment model. Figure 3 This model can compare real-time monitored temperature and strain data with a baseline health state, enabling real-time identification and early warning of abnormal states such as structural hot spots and fatigue cracks, with a temperature measurement error of ≤10% and a strain measurement error of ≤15%.

[0047] 2) Intelligent Diagnostic System Verification: During testing, the host computer receives and processes wavelength drift data from the fiber optic demodulator. Based on the pre-calibrated temperature / strain-wavelength relationship, it calculates the temperature and strain values ​​at each monitoring point in real time. This data is then input into a neural network health assessment model trained using simulation data. In one test, the heat load intensity was artificially increased slightly in a certain area (simulating a hot spot). The system identified an abnormal temperature rise in that area within 3 seconds and issued an early warning. After the test, the accuracy of the warning location was verified by comparing it with infrared images. The temperature measurement error, after calibration, was less than 8%, and the strain measurement error was less than 12%, both meeting the design requirements.

[0048] In summary, this embodiment demonstrates, through a complete preparation and testing process, that the divertor target plate with a membrane rib composite structure proposed in this invention has achieved the expected goals in terms of structural design, manufacturing process, and performance. It can effectively solve the challenges faced by existing technologies and has significant practicality and advancement.

[0049] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A membrane rib composite structure divertor target plate, characterized in that: The target plate has a multi-layer composite structure, which consists of: a plasma-facing material layer, a membrane-ribbed composite heat sink layer, and a backplate support layer, extending inward from the surface that receives the plasma heat flow; the plasma-facing material layer, the membrane-ribbed composite heat sink layer, and the backplate support layer are integrally connected by a stacking forging process.

2. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The plasma-facing material layer is a flat or tile-like structure made of tungsten or tungsten alloy material. The surface facing the plasma is provided with macroscopic or microscopic textures as needed to regulate the interaction between the plasma and the material.

3. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The backplate support layer is a plate-shaped or shell-shaped structure made of stainless steel or high-temperature alloy material, with a pre-set groove inside to accommodate the fiber optic sensor, providing structural support and rigidity for the entire target plate.

4. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The heat sink layer of the membrane rib composite structure is made of copper or copper-chromium-zirconium alloy and has a cooling channel inside. The cooling channel integrates a biomimetic leaf vein rib structure and a vapor-liquid separation membrane, which together constitute a membrane rib composite cooling structure for enhancing heat exchange.

5. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The biomimetic leaf vein rib structure is a turbulence structure set on the inner wall of the cooling channel. Its geometry is inspired by the natural leaf vein network, and it is arranged in a lateral parallel vein structure. That is, the lateral veins branching from the main vein are parallel to each other, and the lateral veins form a certain angle with the main vein. The biomimetic leaf vein rib structure can disturb the fluid in the near-wall region, destroy the laminar boundary layer, increase the vaporization nucleus, and promote subcooled boiling.

6. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The vapor-liquid separation functional membrane is a porous SiO2 membrane disposed on the heated wall surface inside the cooling channel. The porous SiO2 membrane is fixed to the heated wall surface by stacking and forging. Through capillary effect, the liquid coolant continuously wets the heated wall surface, while blocking the bubbles generated by boiling and guiding them to the center of the channel to be carried away by the mainstream, thus avoiding the occurrence of critical heat flux phenomenon.

7. The membrane rib composite structure divertor target plate according to claim 1, characterized in that: The target plate also includes an integrated intelligent diagnostic system, which includes a fiber optic sensor network embedded in the groove of the backplate support layer, a fiber optic demodulator for signal analysis, and a host computer for data processing and health assessment; the fiber optic sensors are embedded in the backplate support layer by an ultrasonic consolidation process.

8. A design method for a divertor target plate based on the membrane rib composite structure of claim 1, comprising: Step 1: Performance Modeling and Analysis Establish a geometric model of the membrane-ribbed composite structure; By observing the parallel vein structure in nature, the main vein structure features are extracted, and then a geometric model is built using 3D CAD software based on the extracted vein structure features. In the model, the biomimetic vein rib structure is arranged with parallel veins extending laterally. The angle between the main vein and the lateral veins is set at 30°~60°, the distance between adjacent lateral veins is 1~5mm, the width of the lateral veins is 1~5mm, and the overall height of the rib is 4~8mm. A macroscopic performance equivalent analysis method for periodic membrane rib units is established by applying the theory of asymptotic homogenization. To accurately analyze stress concentration and potential cracking at the interface of multiple heterogeneous materials in the target plate, interpolation functions for the displacement and temperature fields at the multi-material interface of the target plate element are established based on the extended finite element method: ； in, Let a point be a point within the computational domain; These are standard finite element shape functions; For standard node degrees of freedom, This represents the standard node set; the first term is the standard finite element displacement approximation; the second and third terms are enrichment terms. It is the Heaviside function that describes the displacement discontinuity on both sides of the crack surface. These are the degrees of freedom of the corresponding enriched nodes. It is the set that affects the nodes; It is a family of analytic functions describing the singularity of stress at the crack tip. The term number of this analytic function family, taking values ​​from 1 to 4. These are the degrees of freedom of the corresponding enriched nodes. It is the set that affects the nodes; The damage evolution process of the interface under normal and tangential stresses is characterized by defining a traction-separation constitutive relation; Step 2: Multi-scale optimization of biomimetic design: A multi-objective optimization design model encompassing performance, function, and manufacturing constraints is established. This model aims to maximize heat transfer efficiency and minimize structural stress, while using material usage, fluid pressure drop, and the minimum feature size and maximum overhang angle in the additive manufacturing process as constraints. It is expressed as: ； ； in, Let x be the objective function and x be the design variable. For heat exchange efficiency, For maximum stress, for constraint functions and for ; The non-dominated sorting genetic algorithm NSGA-II was used to solve the multi-objective optimization model.

9. The method according to claim 8, characterized in that: In step 1, the macroscopic performance equivalent analysis method of the periodic membrane rib unit is established using the asymptotic homogenization theory; the specific steps are as follows: a) Extract a representative, periodically repeating minimal structural unit from the geometric model as the representative volumetric unit RVE; b) Apply periodic boundary conditions to the RVE to simulate the stress and heat transfer environment in the macroscopic continuum; c) Solve the heat flux density and stress-strain field response of the RVE under specific boundary conditions using finite element analysis (FEA); d) Calculate the macroscopic equivalent performance parameters using the volume averaging method; assign the properties of the microscopic representative units back to the macroscopic level as the average equivalent values ​​at the macroscopic level; specifically, for any macroscopic physical quantity E and its corresponding microscopic quantity ε, the relationship is: ； in, The operation represents the volume average within the RVE region; the membrane rib composite heat sink layer with complex microstructure is equivalent to a homogeneous material layer with anisotropic macroscopic properties.

10. The method according to claim 8, characterized in that: Step 1 also includes: a) Thermal expansion coefficient mismatch Peak interfacial shear stress They exhibit an approximately linear positive correlation, expressed as ,in, For geometry-related constants, For the equivalent elastic modulus, For working temperature difference; b) Rib height ,width and spacing With equivalent heat transfer coefficient and flow pressure drop This forms a multi-objective optimization surface and is fitted to a response surface function. ; c) The interfacial bonding strength determines the fatigue life of the target plate under thermal cycling; by extending Paris's law, the range of interfacial stress intensity factors is expanded. With crack propagation rate This correlation enables quantitative prediction of thermal cycling fatigue life.

Images