Thermal superstructure manufacturing method, device and system based on discrete transformation thermology

Abstract

The invention discloses a thermal superstructure manufacturing method, device and system based on discrete transformation thermology, and belongs to the technical field of heat conduction and heat flow regulation and control. The method comprises the steps of performing discrete simplex mesh generation on a design domain to obtain a mesh, and appointing a hard constraint for an internal function boundary according to a regulation and control function; piecewise linear mapping is defined on the grid, optimization solving is carried out through two stages, the first stage punishes a negative Jacobian determinant to eliminate folding to obtain bijection mapping, and the second stage minimizes conformal distortion to suppress anisotropy; and deducing a thermal conductivity tensor field of a physical domain by utilizing a Jacobian matrix J, and finally carrying out microstructure matching filling and adopting additive manufacturing. The problems of folding and extreme parameters easily occurring in complex geometric transformation thermal design are solved, efficient thermal management and omni-directional thermal regulation and control under any complex shape can be achieved, and the experimental verification error is smaller than 5%.

Description

Technical Field

[0001] This invention relates to the field of heat conduction and heat flow control technology, and in particular to a method, device and system for manufacturing thermal superstructures based on discrete transformation thermodynamics. Background Technology

[0002] Thermal metastructure devices guide heat flow along a predetermined path through artificial structures, and can be used in scenarios such as thermal protection, thermal management, and thermal field shaping. In the transformational thermal framework, the core idea is to obtain a spatially varying anisotropic thermal conductivity tensor through coordinate transformation, thereby achieving functions such as thermal stealth and heat accumulation.

[0003] Existing techniques often employ analytical function coordinate transformation methods, such as polar coordinate transformation. These methods derive tensors through explicit functions of polar coordinate compression or rotation. However, they have theoretical limitations on geometry, making it difficult to directly cover concave boundaries, multi-connected devices, or highly free-form shapes. Furthermore, analytical transformations can generate extreme anisotropic parameters, leading to material requirements far exceeding the achievable range. Another approach is numerical transformation based on partial differential equations, which generates mappings by solving partial differential equations on a mesh. However, under complex boundary conditions and unstructured meshes, mesh folding easily occurs, causing the Jacobian determinant to become negative and resulting in a non-positive definite thermal conductivity tensor that loses its physical meaning. Moreover, existing methods struggle to simultaneously guarantee fold-free bijective properties and low-distortion manufacturability.

[0004] Therefore, how to obtain bijective coordinate transformations with low distortion on any complex shape has become a technical problem that urgently needs to be solved. Summary of the Invention

[0005] The main objective of this invention is to provide a coordinate transformation that is both bijective and has low distortion, capable of simultaneously achieving a bijective coordinate transformation on any complex shape.

[0006] To achieve the above objectives, this invention proposes a method for manufacturing thermal superstructures based on discrete transformation thermodynamics, comprising the following steps: Step S1: Input the geometric model of the design domain of the thermal superstructure device, and perform discrete simplex meshing on the design domain to obtain the mesh. Specify the mapping semantics for the internal functional boundary according to the target control function and encode it as hard constraint conditions. Step S2: Define a piecewise linear mapping on the grid. and the mapping The solution is transformed into a nonlinear optimization process, wherein: in the first stage, cell folding is eliminated by penalizing the negative Jacobian determinant to obtain a bijective mapping that satisfies the hard constraint condition and whose Jacobian determinant is positive everywhere; in the second stage, the conformal distortion energy is minimized under the iterative framework that preserves bijectivity in order to suppress the anisotropy of the derived tensor. Step S3: Utilize the mapping The Jacobian matrix J(x) at point x, according to the constitutive equation

[0007] Derivation of the corresponding thermal conductivity tensor field in the physical domain ,in The background medium's thermal conductivity; Step S4: Convert the thermal conductivity tensor field Discretize and use a pre-built library of microstructures for matching and filling to generate a multi-scale geometric model; Step S5: Based on the multi-scale geometric model, a superstructure with microstructural features is prepared using additive manufacturing process.

[0008] Preferably, in step S1, the target control function includes at least one of thermal stealth function, thermal heat accumulation function, and thermal rotation function.

[0009] Preferably, in step S1, the thermal stealth function corresponds to the internal functional boundary being compressed to a point region in the virtual space; the thermal heat gathering function corresponds to the internal functional boundary being magnified in the virtual space; and the thermal rotation function corresponds to the internal functional boundary being rotated by a specified angle value in the virtual space.

[0010] Preferably, in step S1, there are two internal functional boundaries, namely a first internal functional boundary and a second internal functional boundary; the first internal functional boundary and the second internal functional boundary are respectively assigned different mapping semantics.

[0011] Preferably, in step S4, the microstructure library is constructed using a periodic homogenization and density-based topology optimization method, with the objective function being to minimize...

[0012] in The single-cell equivalent tensor of the microstructure. Let be the target tensor at the unit to be matched.

[0013] Preferably, in step S5, the superstructure body includes a first phase skeleton; the method further includes filling the pores of the first phase skeleton with a second phase substrate for composite encapsulation.

[0014] Preferably, the first phase skeleton is made of 316L stainless steel, the second phase substrate is polydimethylsiloxane, and the thermal superstructure device further includes thermally conductive paste as a background embedding material.

[0015] Preferably, in step S5, the additive manufacturing process employs laser powder bed melting, and its process parameters include: laser power of 50W, scanning speed of 1000mm / s, and layer thickness of 10. The lane spacing is 50. .

[0016] Preferably, after step S5, a verification step is further included: recording the temperature field of the thermal superstructure device in steady state using an infrared camera, and extracting the normalized root mean square error index on the sampling line. and require it to meet ;in:

[0017] In the formula, This is the normalized value of the experimental temperature. The normalized value for the simulated temperature is given by N, which is the total number of sampling points.

[0018] This application also discloses a thermal superstructure device, which is fabricated using the thermal superstructure fabrication method based on discrete transformation thermodynamics as described in any of the preceding claims.

[0019] This application also discloses a thermal superstructure design system, including: The mesh generation module is used to generate a mesh by discretizing the design domain of thermal superstructure devices using a simplex mesh. The mapping optimization module, whose input is connected to the output of the mesh partitioning module, is used to solve for a piecewise linear mapping on the mesh that satisfies bijective constraints and minimizes conformal distortion energy. ; The tensor derivation module, whose input is connected to the output of the mapping optimization module, is used to utilize the mapping... Derivation of the thermal conductivity tensor field of the physical domain from the Jacobian matrix J ; The matching and filling module, whose input is connected to the output of the tensor derivation module, is used to perform matching and filling on the thermal conductivity tensor field using a pre-built microstructure library. Perform matching and filling to generate a multi-scale geometric model.

[0020] The above technical solution has the following advantages: This invention utilizes discrete simplex meshes to construct piecewise linear mappings, overcoming the limitations of global analytic functions on geometric shapes and enabling transformational thermal design for arbitrarily complex shape domains such as concave boundaries, non-star shapes, and porous structures. By introducing a two-stage decoupled optimization algorithm, the first stage ensures the bijectivity of the mapping by penalizing the negative Jacobian determinant, fundamentally avoiding mesh folding and ensuring the physical validity and positive definiteness of the derived thermal conductivity tensor field. The second stage reduces parameter extrema by minimizing conformal distortion, significantly improving material realizability. Combining multi-scale microstructure matching strategies with additive manufacturing processes, it is possible to accurately realize spatially varying anisotropic tensor distributions, resulting in stable performance of the fabricated thermal superstructure devices under omnidirectional incidence conditions. The normalized root mean square error of the temperature profile in experiments and simulations can be stably controlled within 5%, providing a general solution for efficient thermal management in complex geometric environments. Attached Figure Description

[0021] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings, wherein: Figure 1 The prior art and the method of the present invention are compared and illustrated with an end-to-end process diagram provided for the embodiments of the present invention.

[0022] in, Figure 1 (a) shows the basic categories and characteristics of thermal operation methods; the M region represents the operation region. “BT” represents the background temperature, Mat1 and Mat2 represent two different base materials, R1-R4 are different regions, and the indices are from 1 to 4. Figure 1 Figures (b)-(d) provide a brief overview of transformational thermal strategies based on different geometric mapping methods: Figure 1 In the diagram, (b) represents the polar coordinate transformation and analytical transformation function; Figure 1 (c) in the figure represents the Laplace transform over a regular grid; Figure 1 In the diagram, (d) represents the Laplace transform over an irregular triangular mesh. Figure 1 (c) and Figure 1 The purple element in (d) represents the region that folds after the transformation.

[0023] Figure 2 This is an end-to-end design flowchart of discrete transformation thermodynamics provided in an embodiment of the present invention. Figure 2 (a) shows the geometry of the thermal cloak, thermal concentrator, and thermal rotator. Figure 2 (b)-(e) are simplified illustrations of the mesh transformation strategy provided in the embodiments of the present invention: Figure 2 (b) in the diagram represents the initial mesh generation; Figure 2 (c) in the figure is a sketch of the internal boundary of the target area; Figure 2In this context, (d) represents the solution for a bijective mapping. Figure 2 In the equation (e), the goal is to minimize conformal distortion to obtain a quasi-conformal mapping. Figure 2 In the figure, (f) represents the thermal conductivity tensor distribution of the three metamaterial devices. Figure 2 The (g)-(h) in the model are based on the thermal conductivity tensor distribution and are filled by a two-dimensional microstructure library to generate multi-scale metamaterial devices.

[0024] Figure 3 The simulation and omnidirectional performance diagram of the two-dimensional free-form device provided in the embodiments of the present invention. Figure 3 Simulation results for three thermal metamaterial devices are presented. Figure 3 (I) in the figure represents the mesh discretization diagram of the thermal metamaterial device before and after transformation; Figure 3 (II) in the figure represents the simulation results under the normalized scale; Figure 3 (III) in the figure is the mesh quality analysis, comparing the results obtained by solving the PDE (left figure) with the results obtained by this method (right figure); Figure 3 (IV) represents the thermal conductivity tensor distribution of multiscale thermal metamaterial devices and their internal filling microstructures; Figure 3 (V) in the figure represents the full-scale simulation result; Figure 3 (VI) in the figure represents the omnidirectional performance analysis of multi-scale thermal metamaterial devices under different directions of heat flow. The polar coordinate plot shows the thermal rotation and invisibility cloak effect in all directions.

[0025] Figure 4 Simulation and index diagram of a three-dimensional free-form heat concentrator provided for embodiments of the present invention. Figure 4 This demonstrates the application of the invention in a three-dimensional context. Figure 4 (a) shows the geometry of a three-dimensional thermal metamaterial device, with a Stanford Rabbit model inserted into the cloud; Figure 4 (b) in the figure is a transformation diagram of the heat concentration function, which shows that the method effectively realizes the heat concentration function in a three-dimensional context, while the Laplace mesh transformation fails. Figure 4 (c) in the figure is a distribution diagram of different components of the thermal conductivity tensor; Figure 4 (d)-(f) in the figure represent the heat transfer simulation of the heat concentrator under the action of heat flow along the x-axis, y-axis and z-axis; Figure 4 (g)-(h) in the diagram are schematic diagrams of covering an irregular tetrahedral mesh on a 20×20×20 regular mesh and optimizing the thermal conductivity tensor of the microstructure using OpenTM. Figure 4 In the diagram, (i) represents the final three-dimensional multi-scale thermal concentrator.

[0026] Figure 5 Example diagram of a multifunctional device provided in an embodiment of the present invention. Figure 5 Two example devices were shown. Figure 5(a) in the image represents a cat and dog metamaterial device. Figure 5 (b) in the image represents a Jinli metamaterial device. Figure 5 In the diagram, (I) represents the discretized grid before and after the transformation. Figure 5 (II) in the figure represents the simulation results under the normalized scale; Figure 5 (III) in the figure is a mesh quality analysis, which shows a comparison between solving PDE (left figure) and the present method (right figure); Figure 5 (IV) represents the thermal conductivity tensor distribution of multi-scale metamaterial devices and their internal filling microstructures; Figure 5 (V) in the figure represents the full-scale simulation result; Figure 5 (VI) in the figure represents the omnidirectional performance analysis of the multi-scale thermal metamaterial device under different angles of heat flow. The polar coordinate plot shows the thermal rotation of the cat-shaped region and the thermal cloak effect of the dog-shaped region.

[0027] Figure 6 The printed sample and experimental verification diagram provided for embodiments of the present invention. Figure 6 This showcases metal parts obtained through 3D printing. Figure 6 (a)-(c) are printed samples of the thermal invisibility cloak, thermal concentrator and thermal rotator (using SS316 stainless steel). Figure 6 (d)-(f) are the corresponding local CAD models of the thermal cloak, thermal concentrator, and thermal rotator; Figure 6 (g)-(i) in the figure represent the actual metamaterial device after the hot paste has been injected; Figure 6 (j) in the diagram represents the experimental testing scenario, demonstrating the experimental schemes for applying high / low temperatures at room temperature to the left and right edges of the metamaterial device. A magnified view of the partial image is shown below. Figure 6 (k) in; Figure 6 In the figure, (l) represents the experimental temperature distribution of the thermal invisibility cloak under the action of heat flow in the x-axis direction; Figure 6 In the figure, (m) represents the experimental temperature distribution of the thermal invisibility cloak under the action of heat flow in the y-axis direction; Figure 6 In the equation (n)-(o), the result is related to the simulation results. Figure 6 (l) and Figure 6 Comparison on the temperature sampling line (m) in the middle; Figure 6 (p)-(s) and Figure 6 In the figure, (t)-(w) represent experimental verifications on a heat concentrator and a heat rotator, respectively, and the temperature distribution data have been standardized. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the following description is provided in conjunction with the appendix. Figures 1 to 6The present invention will be described in detail with reference to the embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. In the following description, specific details, such as particular system architectures, techniques, etc., are set forth for explanation and not limitation in order to provide a thorough understanding of the embodiments of the present invention. However, those skilled in the art will recognize that the present invention may be implemented in other embodiments without these specific details.

[0029] Example 1 like Figure 1 and Figure 2 As shown, this embodiment provides a method for designing and manufacturing two-dimensional arbitrary-shape thermal superstructure devices based on discrete transformation thermodynamics. This method aims to address the problems in existing technologies, such as the strict limitations on geometry imposed by analytical transformation methods, and the tendency of numerical grid transformation methods to suffer from element folding (i.e., negative Jacobian determinants) when dealing with complex boundaries, leading to the loss of physical meaning in the derived thermal conductivity tensor.

[0030] First, step S1 is executed to perform geometric input and functional encoding. In this embodiment, the geometric model of the design domain of the thermal superstructure device is input. The outer boundary of this design domain can be an arbitrarily complex closed curve, such as a free shape with multiple concave regions, including concave boundaries, non-star-shaped or multi-connected hole structures. An internal functional boundary is defined within the design domain; the area enclosed by this internal functional boundary is the core area where specific thermal control functions need to be implemented. For this internal functional boundary, a mapping semantic is specified according to the preset target control function. For example, if the target control function is thermal stealth, the internal functional boundary is mapped in virtual space to a very small region, i.e., an ideal convergent point region, so that external heat flux propagates around this region in physical space, achieving thermal shielding of the core area. If the target control function is thermal accumulation, the internal functional boundary is correspondingly magnified and mapped in virtual space to enhance the heat flux or temperature gradient of the core area. If the target control function is thermal rotation, a specific angle value is specified for the internal functional boundary to rotate relative to physical space in virtual space, for example… This is to redirect the direction of heat flow. These mapping semantics are encoded as hard constraints in the discrete mapping solution process, namely Dirichlet boundary conditions, which are expressed as follows: , where the target geometry Defined by the functional constraint function.

[0031] Subsequently, the design domain is discretized into a simplex mesh to obtain the mesh. In this embodiment, an unstructured triangular mesh is used to fill the two-dimensional planar domain to ensure that the mesh completely covers the entire design domain, and appropriate densification is applied at the boundaries according to the geometric complexity to ensure the accuracy of subsequent optimization calculations.

[0032] Execute step S2 to solve and optimize the discrete mapping. Define a piecewise linear mapping on the mesh. .

[0033] This method eliminates mesh folding by minimizing the following global energy functional and ensures that the positions of the outer and inner boundary vertices remain fixed:

[0034] in, It is a vector of all vertex variables. It is a simplex The volume (or can be simply set to 1). Energy term. and The definition is as follows:

[0035]

[0036] in, It is a smooth determinant regularization operator:

[0037] By implementing a continuation strategy, The objective function acts as a barrier, ensuring that all units are pushed to the effective configuration (i.e., This yields a locally injective, unfold-free mapping. Since the specified boundary mapping is non-self-intersecting, the final mapping is bijective.

[0038] Once an effective mesh has been established using a foldless algorithm, conformal smoothing is then performed to further optimize the quality of the mapping. For this purpose, the present invention utilizes the Most Isometric Parameterization (MIPS) energy, which provides a robust measure of conformality.

[0039] For each individual triangle, the MIPS energy is defined as the ratio of the Frobenius norm of the Jacobian matrix to the square of its determinant:

[0040] in, It is spatial dimension (in the case of two dimensions) In three dimensions This energy function penalizes non-conformal distortions while maintaining scale invariance.

[0041] The overall goal is:

[0042] in, The weights can be regionally weighted or uniformly weighted. This nonlinear optimization is solved using interior-point methods and element-wise line search; the gradient is calculated analytically to ensure robustness. The MIPS step effectively reduces the condition number of all Jacobian matrices, improving the regularity of the thermal conductivity tensor field.

[0043] Through the aforementioned two-stage optimization algorithm and conformal smoothing process, this invention can optimize the geometric transformation of the thermal superstructure while ensuring physical feasibility. The foldless mapping algorithm ensures that there are no unreasonable folds in the design of the thermal superstructure, while conformal smoothing further improves the continuity and uniformity of the thermal conductivity tensor, ensuring that the final design has stable and efficient thermal performance.

[0044] Step S3 is executed to derive the thermal conductivity tensor field. The optimized piecewise linear mapping is then used. Jacobian matrix at grid cell point x Based on the constitutive equations of transformation thermodynamics, the corresponding thermal conductivity tensor field in the physical domain is derived. The specific calculation formula is as follows:

[0045] in, The thermal conductivity of the background medium is preset and, in this embodiment, is set to a specific isotropic scalar based on the experimental environment. This step transforms the complex geometric spatial transformation into a non-uniform, anisotropic thermal conductivity distribution in physical space. Since the aforementioned optimization process ensures that the Jacobian determinant is positive, the resulting thermal conductivity tensor for each unit is... All are symmetric and positive definite tensors, satisfying the physical feasibility requirement.

[0046] Execute step S4 to generate a multi-scale geometric model. The derived continuous thermal conductivity tensor field... Discretize to a predefined topological functional unit, for example, divide the device domain into... Each functional unit is filled in unit by unit using a pre-built microstructure library. The microstructure library is pre-built using a periodic homogenization and density-based topology optimization method, i.e., the SIMP method. The periodic homogenization process satisfies the volume average relation of the local problem in a single cell. ,in Let be the equivalent thermal conductivity tensor of the microstructure. During the construction process, the optimization objective function is set to minimize the unit cell equivalent tensor of the microstructure. Target tensor at the unit to be matched The norm difference between them can be expressed mathematically as:

[0047] Simultaneously, volume fraction, filtering, regularization, and connectivity constraints are applied during the optimization process. Through this matching strategy, the macroscopic tensor field distribution is transformed into a multi-scale geometric model with specific microstructural features. Each functional unit is filled with a metal skeleton structure capable of achieving the target anisotropic thermal conductivity. For example... Figure 3 As shown, based on the simulation comparison results, the relative 2-norm error of the temperature field of the thermal cloak in this embodiment between the homogenized model and the full-scale microstructure simulation is approximately... The difference between a thermal concentrator and a thermal rotator is approximately... to between.

[0048] Step S5 is executed to fabricate the superstructure device. Based on the generated multi-scale geometric model, a superstructure substrate with microstructural features is fabricated using laser powder bed melting (LPBF) additive manufacturing. In this embodiment, the manufacturing equipment employs a continuous-wave fiber laser system with a laser wavelength of [wavelength missing]. The diameter of the focused spot is The superstructure consists of gas atomization. The first-phase framework is made of stainless steel powder, with a powder particle size distribution of... to Between. The parameters for the LPBF process are set as follows: laser power is The scanning speed is The layer thickness is Lane spacing is Interlayer adopts Scanning and rotation are employed to reduce thermal distortion. The manufacturing environment is protected by high-purity nitrogen, with oxygen content controlled to less than [value missing]. This high-precision manufacturing method ensures the dimensional accuracy of the micro-framework structure, thereby guaranteeing the accurate realization of thermal properties. After the first-phase framework is prepared, a second-phase substrate is filled into the pores of the first-phase framework to form a complete composite system and reduce contact thermal resistance. In this embodiment, polydimethylsiloxane (PDMS) is used as the second-phase substrate for composite encapsulation. After the superstructure body is prepared, thermally conductive adhesive is poured into a mold as a background material and plasticized into a square shape for subsequent experimental measurements.

[0049] like Figure 6 As shown, the final verification was performed. An infrared camera was used to record the temperature field of the thermal superstructure device in steady state. A constant high-temperature boundary was applied to both sides of the device. With low temperature boundary This forms a stable heat flow. By extracting temperature data from the sampling line, it is first normalized according to the boundary temperature difference. The normalized temperature formula is as follows:

[0050] The normalized experimental temperature value was then calculated. With simulated temperature value The normalized root mean square error index (NRMSE) is used to measure the error between the two values. The formula for calculating NRMSE is as follows:

[0051] Where N is the total number of sampling points. Experimental results show that the device fabricated in this embodiment exhibits extremely high performance robustness in omnidirectional scanning mode. Specifically, for the thermal stealth cloak, the average thermal flux penetration ratio of the core region is only [value missing]. The maximum value is Corresponding angle of incidence The minimum value is Corresponding angle of incidence For a thermal concentrator, its average heat concentration ratio is: The maximum value is Corresponding angle of incidence The minimum value is Corresponding angle of incidence For a thermal rotator, the average heat flow deflection angle is... The maximum value is Corresponding angle of incidence The minimum value is Corresponding angle of incidence The sampling line NRMSE of all devices under different loading directions is less than [value missing]. This demonstrates that the discrete transformation thermodynamics-based method proposed in this invention can accurately design and manufacture superstructure devices that meet complex thermal control requirements, and has extremely high predictive consistency.

[0052] Example 2 like Figure 4 As shown, this embodiment provides a design and simulation verification method for a three-dimensional arbitrary-shape heat concentrator. In this embodiment, the outer boundary of the constructed device is a free-form three-dimensional cloud shape, and its internal functional boundary is defined as an arbitrary three-dimensional shape region, specifically a bunny-shaped functional area, aiming to achieve heat concentration effect in three-dimensional space and enhance the average heat flux of the core region. The specific implementation steps are as follows: First, the entire design domain is discretized using a tetrahedral mesh. Based on the target function, i.e., the heat concentration function, hard constraints are encoded on the internal functional boundary, and then solved using the aforementioned two-stage decoupling optimization algorithm. In the first stage, the negative Jacobian determinant is penalized to eliminate cell folding and obtain a bijective mapping, ensuring that the Jacobian determinant satisfies... In the second stage, conformal distortion energy is minimized to reduce the derived anisotropy. This is achieved using mapping. The Jacobian matrix J is used to derive the three-dimensional thermal conductivity tensor field based on the constitutive equation. The derived tensor at this point contains diagonal terms. , , and off-diagonal terms , , .

[0053] In the multi-scale implementation process, this embodiment adopts A regular mesh covers the unstructured tetrahedral domain, and each cell is assigned a corresponding target tensor value before microstructure matching. Regarding the material system, the first phase framework of the three-dimensional microstructure is selected from a high thermal conductivity material, such as aluminum alloy. The second phase substrate uses a low thermal conductivity material, such as polydimethylsiloxane (PDMS), for composite filling. Simulation results show that under three orthogonal loading conditions (x-axis loading, y-axis loading, and z-axis loading), the ratio of the average heat flux amplitude in the core functional area, i.e., the measured heat flux ratio, performs well. Specifically, the average heat flux ratio under x-axis loading is... The average heat ratio under y-direction loading is The average heat ratio under z-axis loading is The above data demonstrate that the method of the present invention can still provide stable and efficient thermal control performance in three-dimensional freeform devices.

[0054] Example 3 like Figure 5 As shown, this embodiment provides an implementation scheme for a multifunctional integrated thermal superstructure device. Two independent internal functional boundaries are defined in the design domain: a first internal functional boundary and a second internal functional boundary. Different mapping semantics are assigned to these two boundaries; for example, a thermal rotation semantic is assigned to the first internal functional boundary, requiring the core region heat flow to rotate. The second internal functional boundary is assigned thermal stealth semantics, requiring thermal shielding of the core area. Through a unified discrete simplex mesh generation and optimization process, multiple functions are integrated within a single device. During optimization, the hard constraints of both internal boundaries are applied simultaneously, ensuring overall consistency and omnidirectional robustness of the mapping. This scheme demonstrates the flexibility of the invention in handling multi-objective control tasks, enabling the integration of multiple functions such as thermal stealth, heat accumulation, and thermal rotation within a single device according to actual needs.

[0055] Alternative implementation methods In addition to the embodiments described above, this invention also supports various variant schemes. Regarding the optimizer and distortion energy form, distortion energy... Different forms of conformal or quasi-conformal distortion metrics can be used. The optimizer in stage two can be replaced with interior-point methods, barrier function methods, or other second-order or quasi-Newton optimizers, as long as they can continuously reduce distortion and maintain bijectivity while preserving hard constraints. Regarding mesh and discrete resolution, the mesh resolution, for example, is two-dimensional... Unit and three-dimensional The unit is merely an example and can be adjusted based on the minimum feature size required for manufacturing, computational resources, and target accuracy. Regarding material systems, the combination of high thermal conductivity and low thermal conductivity phases is not limited to... Stainless steel or aluminum alloys and polydimethylsiloxane can be replaced with any system that meets the requirements for thermal conductivity difference, such as metal and air, metal and vacuum pores, or metal and various low thermal conductivity materials (such as thermal paste).

[0056] Regarding manufacturing and packaging processes, the filling process can select different potting compounds or thermally conductive interface materials according to actual needs to further reduce the contact thermal resistance between the microstructure and the substrate. Furthermore, in the verification phase, the steady-state temperature field is recorded using an infrared camera, and the normalized root mean square error index (NRMSE) is calculated. The calculation formula is as follows: Experimental results show that the sampling line NRMSE of each device can be stably controlled under different loading directions. Within. By modifying the semantics of boundary constraints and the combination of internal regions, the method of this invention can also be extended to the design and manufacture of various thermal field shaping devices such as thermal deflection, thermal conduction, and thermal shunting.

[0057] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for manufacturing thermal superstructures based on discrete transformation thermodynamics, characterized in that, Includes the following steps: Step S1: Input the geometric model of the design domain of the thermal superstructure device, and perform discrete simplex meshing on the design domain to obtain the mesh. Specify the mapping semantics for the internal functional boundary according to the target control function and encode it as hard constraint conditions. Step S2: Define a piecewise linear mapping on the grid. and the mapping The solution is transformed into a nonlinear optimization process, wherein: in the first stage, cell folding is eliminated by penalizing the negative Jacobian determinant to obtain a bijective mapping that satisfies the hard constraint condition and whose Jacobian determinant is positive everywhere; in the second stage, the conformal distortion energy is minimized under the iterative framework that preserves bijectivity in order to suppress the anisotropy of the derived tensor. Step S3: Utilize the mapping The Jacobian matrix J(x) at point x, according to the constitutive equation Derivation of the corresponding thermal conductivity tensor field in the physical domain ,in The background medium's thermal conductivity; Step S4: Convert the thermal conductivity tensor field Discretize and use a pre-built library of microstructures for matching and filling to generate a multi-scale geometric model; Step S5: Based on the multi-scale geometric model, a superstructure with microstructure features is prepared using additive manufacturing process.

2. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, In step S1, the target control function includes at least one of thermal stealth function, thermal heat accumulation function, and thermal rotation function.

3. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 2, characterized in that, In step S1, the thermal stealth function corresponds to the internal functional boundary being compressed to a point region in the virtual space; the thermal heat concentration function corresponds to the internal functional boundary being magnified in the virtual space. The hot rotation function corresponds to the internal functional boundary rotating by a specified angle value in virtual space.

4. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, In step S1, there are two internal functional boundaries, namely a first internal functional boundary and a second internal functional boundary; the first internal functional boundary and the second internal functional boundary are respectively assigned different mapping semantics.

5. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, In step S4, the microstructure library is constructed using a periodic homogenization and density-based topology optimization method, with the objective function being to minimize... in The single-cell equivalent tensor of the microstructure. Let be the target tensor at the unit to be matched.

6. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, In step S5, the superstructure body includes a first phase skeleton; the method further includes filling the pores of the first phase skeleton with a second phase substrate for composite encapsulation.

7. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 6, characterized in that, The first phase skeleton is made of 316L stainless steel, the second phase substrate is polydimethylsiloxane, and the thermal superstructure device also includes thermally conductive paste as a background embedding material.

8. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, In step S5, the additive manufacturing process employs laser powder bed melting, with process parameters including: laser power of 50W, scanning speed of 1000mm / s, and layer thickness of 10. The lane spacing is 50. .

9. The method for manufacturing thermal superstructures based on discrete transformation thermodynamics according to claim 1, characterized in that, Following step S5, a verification step is also included: recording the temperature field of the thermal superstructure device under steady state using an infrared camera, and extracting the normalized root mean square error index from the sampling line. and require it to meet ;in: In the formula, This is the normalized value of the experimental temperature. The normalized value for the simulated temperature is given by N, which is the total number of sampling points.

10. A thermal superstructure device, characterized in that, The thermal superstructure was fabricated using the method for manufacturing thermal superstructures based on discrete transformation thermodynamics as described in any one of claims 1 to 9.

11. A thermal superstructure design system, characterized in that, include: The mesh generation module is used to generate a mesh by discretizing the design domain of thermal superstructure devices using a simplex mesh. The mapping optimization module, whose input is connected to the output of the mesh partitioning module, is used to solve for a piecewise linear mapping on the mesh that satisfies bijective constraints and minimizes conformal distortion energy. ; The tensor derivation module, whose input is connected to the output of the mapping optimization module, is used to utilize the mapping... Derivation of the thermal conductivity tensor field of the physical domain from the Jacobian matrix J ; The matching and filling module, whose input is connected to the output of the tensor derivation module, is used to perform matching and filling on the thermal conductivity tensor field using a pre-built microstructure library. Perform matching and filling to generate a multi-scale geometric model.

Images