Simulation data set construction method based on silicon carbide chip heat dissipation structure design

Abstract

The invention discloses a simulation data set construction method based on silicon carbide chip heat dissipation structure design, and particularly relates to the technical field of chip heat dissipation, and the method comprises the steps: firstly constructing a heat dissipation structure design database, collecting data needed by heat dissipation structure design simulation, and determining various parameter combinations and corresponding experimental data; the method comprises the following steps: constructing a simulation model of a silicon carbide chip heat dissipation structure, constructing a multi-dimensional simulation data set through the constructed simulation model of the silicon carbide chip heat dissipation structure, processing the multi-dimensional simulation data set, constructing a heat dissipation structure design simulation data set according to the processed simulation data set, feeding back an abnormal processing result to a man-machine end for analysis and evaluation, and carrying out man-machine interaction according to an abnormal evaluation result; according to the method, data analysis and mining are carried out based on the data set, the internal relation between the data generated in the design process of the heat dissipation structure and the thermal performance can be found, a design administrator is helped to formulate a more scientific optimization strategy, and the heat dissipation efficiency and the performance stability of the heat dissipation structure are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of chip heat dissipation, and specifically to a method for constructing a simulation data set based on the heat dissipation structure design of silicon carbide chips. Background Technique

[0002] Due to its excellent high-temperature, high-voltage, and high-frequency characteristics, silicon carbide chips are widely used in the fields of power electronics, new energy vehicles, etc. However, a large amount of heat is generated during the operation of silicon carbide chips. Therefore, an efficient heat dissipation structure design is crucial for the normal operation of silicon carbide chips. During the heat dissipation structure design process, simulation technology is an important auxiliary means. Through simulation, the performance of the heat dissipation structure can be predicted in the design stage, reducing the number of experiments and lowering the R & D cost and cycle.

[0003] In terms of the heat dissipation structure design of silicon carbide chips, traditional methods mainly rely on empirical design and repeated experiments. Engineers initially determine the form and parameters of the heat dissipation structure based on past design experience, and then verify the rationality of the design by actually manufacturing samples and conducting thermal tests. With the development of computer technology, numerical simulation technology has gradually been applied to the heat dissipation structure design of chips. Existing simulation models often rely on simplified assumptions and idealized conditions, and there is a certain gap from the actual heat dissipation situation of silicon carbide chips.

[0004] A high-quality simulation data set is the basis for accurate simulation. Currently, the existing simulation models still have the following problems: on the one hand, empirical design is not only time-consuming and laborious, increasing the R & D cost, but also due to the limitations of actual test conditions, it is difficult to comprehensively obtain detailed thermal performance data of the heat dissipation structure under various working conditions; on the other hand, the data sets constructed by existing methods rely on single-parameter experiments and lack multi-parameter coupling analysis. Therefore, the existing heat dissipation structure design methods for silicon carbide chips have deficiencies in terms of accuracy, efficiency, and data support, and there is an urgent need for an innovative method for constructing a simulation data set based on the heat dissipation structure design of silicon carbide chips to improve the scientificity and effectiveness of heat dissipation structure design. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for constructing a simulation data set based on the heat dissipation structure design of silicon carbide chips to solve the problems raised in the above background technique.

[0006] To achieve the above object, the present invention provides the following technical solution: A method for constructing a simulation data set based on the heat dissipation structure design of silicon carbide chips, including: S1: Through big data technology, collect, store, and update the influencing factors during the heat dissipation process of silicon carbide chips, and receive the data generated in the heat dissipation structure design of silicon carbide chips to construct a heat dissipation structure design database; S2: Based on the heat dissipation structure design database obtained in S1, collect the data required for the simulation of the heat dissipation structure design, and determine various parameter combinations and corresponding experimental data for constructing the simulation dataset; S3: Through the simulation model construction technology, construct a simulation model of the silicon carbide chip heat dissipation structure based on the first text of the heat dissipation structure design and the second text of the heat dissipation structure design obtained in S2; S4: Input the various parameter combinations obtained in S2 into the simulation model of the silicon carbide chip heat dissipation structure obtained in S3 to construct a multi-dimensional simulation dataset, including the chip geometric parameter and heat dissipation performance dataset, the heat dissipation structure parameter and heat dissipation efficiency dataset, the material property and thermal interface performance dataset, and the working condition and dynamic heat dissipation response dataset; S5: Through the big data processing technology, process the multi-dimensional simulation dataset obtained in S4 to obtain the heat dissipation structure design simulation dataset after data processing, including the processed passing results and the processed abnormal results, and feedback the processed abnormal results to the design administrator; S6: Through the big data analysis technology, evaluate the processed abnormal results, and feedback to the design administrator according to the abnormal evaluation results for human-computer interaction.

[0007] Technical effects and advantages of the present invention: 1. By comprehensively analyzing the key factors affecting the performance of the silicon carbide chip heat dissipation structure, reasonably determining the parameter value range, and using the orthogonal experimental design to generate four types of parameter combinations, the present invention can cover a wide range of actual working conditions with fewer samples, ensuring that the constructed simulation dataset has good comprehensiveness and representativeness, and making the simulation results more in line with the actual situation; 2. Based on the dataset for data analysis and mining, the present invention can discover the internal relationship and law between the data generated in the heat dissipation result design process and the thermal performance, help the design administrator formulate a more scientific optimization strategy, improve the heat dissipation efficiency and performance stability of the heat dissipation structure; without the need for a large number of actual tests, thus greatly shortening the R & D cycle of the silicon carbide chip heat dissipation structure; 3. Through the quality evaluation and feedback adjustment mechanism for the dataset, the present invention can timely discover and solve the problems existing in the dataset, further improve the quality of the dataset, meet the requirements for simulation data under different application scenarios, help improve the efficiency and reliability of the silicon carbide chip heat dissipation structure design, and reduce the R & D cost and cycle. Brief Description of the Drawings

[0008] Figure 1 It is a schematic diagram of the overall process of the present invention.

[0009] Figure 2 It is a schematic diagram of the method process of the present invention. Detailed Embodiment

[0010] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0011] Please refer to Figure 1 As shown, the present invention provides a simulation data set construction system based on the heat dissipation structure design of a silicon carbide chip, including a heat dissipation structure design data management center, a heat dissipation structure design simulation data set determination module, a heat dissipation structure design simulation model construction module, a heat dissipation structure design simulation data generation module, a heat dissipation structure design simulation data set processing and construction module, and a heat dissipation structure design simulation data set construction human-machine module.

[0012] The heat dissipation structure design data management center is connected to the other modules. The heat dissipation structure design simulation data set determination module is respectively connected to the heat dissipation structure design simulation model construction module and the heat dissipation structure design simulation data generation module. The heat dissipation structure design simulation data generation module is respectively connected to the heat dissipation structure design simulation model construction module and the heat dissipation structure design simulation data set processing and construction module. The heat dissipation structure design simulation data set construction human-machine module is connected to the heat dissipation structure design simulation data set processing and construction module.

[0013] Heat dissipation structure design data management center: Through big data technology, collect, store, and update the influencing factors in the heat dissipation process of the silicon carbide chip, and receive the data generated in the heat dissipation structure design of the silicon carbide chip to construct a heat dissipation structure design database; Heat dissipation structure design simulation data set determination module: Based on the heat dissipation structure design database, collect the data required for heat dissipation structure design simulation, determine various parameter combinations for constructing the heat dissipation structure and the corresponding experimental data, and transmit them to the heat dissipation structure design simulation data generation module; Heat dissipation structure design simulation model construction module: Through simulation model construction technology, construct a simulation model of the heat dissipation structure of the silicon carbide chip according to the data required for heat dissipation structure design simulation, and transmit it to the heat dissipation structure design simulation data generation module; Heat dissipation structure design simulation data generation module: Used to input various parameter combinations into the constructed simulation model of the heat dissipation structure of the silicon carbide chip, construct a multi-dimensional simulation data set, and transmit it to the heat dissipation structure design simulation data set processing and construction module; Heat dissipation structure design simulation data set processing and construction module: Through big data processing technology, process multi-dimensional simulation data sets, construct a heat dissipation structure design simulation data set based on the passed simulation data sets, and at the same time feedback the abnormal processing results to the heat dissipation structure design simulation data set construction human-machine module; Heat dissipation structure design simulation data set construction human-machine module: Through big data analysis technology, evaluate the abnormal processing results and perform human-computer interaction based on the evaluated abnormal results.

[0014] Please refer to Figure 2 As shown, a method for constructing a simulation data set for the heat dissipation structure design of a silicon carbide chip includes: S1: Through big data technology, collect, store, and update the influencing factors during the heat dissipation process of the silicon carbide chip, and receive the data generated in the heat dissipation structure design of the silicon carbide chip to construct a heat dissipation structure design database; S2: Based on the heat dissipation structure design database obtained in S1, collect the data required for heat dissipation structure design simulation, and determine various parameter combinations and corresponding experimental data for constructing the simulation data set; S3: Through simulation model construction technology, based on the first text of the heat dissipation structure design and the second text of the heat dissipation structure design obtained in S2, construct a simulation model of the heat dissipation structure of the silicon carbide chip; S4: Input the various parameter combinations obtained in S2 into the simulation model of the heat dissipation structure of the silicon carbide chip obtained in S3 to construct a multi-dimensional simulation data set, including a chip geometry parameter and heat dissipation performance data set, a heat dissipation structure parameter and heat dissipation efficiency data set, a material property and thermal interface performance data set, and a working condition and dynamic heat dissipation response data set; S5: Through big data processing technology, process the multi-dimensional simulation data set obtained in S4 to obtain a heat dissipation structure design simulation data set after data processing, including passed processing results and abnormal processing results, and feedback the abnormal processing results to the design administrator; S6: Through big data analysis technology, evaluate the abnormal processing results, and feedback to the design administrator according to the abnormal evaluation results for human-computer interaction.

[0015] S1: Through big data technology, collect, store, and update the influencing factors during the heat dissipation process of the silicon carbide chip, and receive the data generated in the heat dissipation structure design of the silicon carbide chip to construct a heat dissipation structure design database; In this embodiment, it should be specifically noted that through industry packages, consulting a large number of academic literatures, conducting preliminary experimental studies, etc., the influencing factors in the heat dissipation process of silicon carbide chips are sorted out; starting from the characteristics of the chips themselves, considering the influence of the geometric dimensions of the chips on the thermal conductivity and specific heat capacity, and then determining the geometric parameters and material property parameters of the chips; for the heat dissipation structure, analyzing the role of the material and thickness of the heat dissipation substrate in heat conduction, and determining the geometric parameters and material parameters of the heat dissipation structure. At the same time, various conditions in the actual operation of the chips are considered, such as working condition parameters such as different power input levels, ambient temperature change ranges, and flow rates and temperatures of cooling media (such as air, liquid coolants).

[0016] S2: Based on the heat dissipation structure design database obtained in S1, collect the data required for the heat dissipation structure design simulation, and determine various parameter combinations and corresponding experimental data for constructing the simulation data set, including the following steps: S2.1: Based on the heat dissipation structure design database obtained in S1, collect the data required for the heat dissipation structure design simulation, and obtain the first text of the heat dissipation structure design, including the length, width, and thickness of the chip; the height, spacing, and shape of the heat dissipation fins, the thickness and area of the heat dissipation substrate; the thermal conductivity and specific heat capacity of the silicon carbide chip, the thermal conductivity and emissivity of the heat dissipation structure material (such as ordinary silicone grease, phase change material, liquid metal, etc.), and the thermal conductivity and interfacial thermal resistance of the filling material between the chip and the heat dissipation structure; the input power of the chip, ambient temperature, flow rate and temperature of the cooling medium, etc.; S2.2: Based on industry standards, determine the ranges of the data required for the heat dissipation structure design simulation obtained in S2.1, and obtain the second text of the heat dissipation structure design, including the determination of the chip geometric parameter range, the heat dissipation structure geometric parameter range, the material property parameter range, and the working condition parameter range; In this embodiment, it should be specifically noted that for the determination of the parameter range, for example, referring to common silicon carbide chip products on the market, the length and width parameter ranges of the chip size are set to 1 mm to 10 mm, and the thickness range is set to 0.1 mm to 1 mm; for the heat dissipation fins, the fin height range is set to 5 mm to 50 mm, the fin spacing range is set to 1 mm to 5 mm, and the fin shape can be selected from common shapes such as rectangular and needle-shaped; for the heat dissipation substrate, the thickness range is set to 1 mm to 10 mm, and the area is appropriately enlarged according to the chip size; in terms of material properties, according to the performance indexes of existing materials, the thermal conductivity range of the silicon carbide chip is set to 300 W / (m·K) to 500 W / (m·K), the thermal conductivity range of the heat dissipation structure material (such as copper, aluminum, etc.) is set to 200 W / (m·K) to 400 W / (m·K), the thermal conductivity range of the filling material between the chip and the heat dissipation structure is set to 1 W / (m·K) to 10 W / (m·K), and the interfacial thermal resistance range is 1×10 -4From \(m^2\cdot K / W\) to \(1\times10^{-3}m^2\cdot K / W\). In terms of working conditions, the chip input power range is set from 1 W to 100 W, the ambient temperature range is set from 20 °C to 80 °C. For air cooling, the cooling air flow rate range is set from 1 m / s to 10 m / s; for liquid cooling, the coolant temperature range is set from 10 °C to 50 °C, and the flow rate range is set from 0.1 m / s to 1 m / s.

[0017] S2.3: First, according to the determined parameter ranges obtained in S2.2, select four types of parameter combinations to generate the third text of the heat dissipation structure design, including the chip geometric parameter combination, the heat dissipation structure parameter combination, the material property parameter combination, and the working condition parameter combination; among them, the chip geometric parameter combination includes the length, width, and thickness of the chip, the thermal conductivity of the chip material, and the convective heat transfer coefficient; the heat dissipation structure parameter combination includes the convective heat transfer coefficient, fin height, fin pitch, fin thickness, and the thermal conductivity of the fin material; the material property parameter combination includes the thermal conductivity of silicon carbide and the thermal conductivity of the filling material, the contact area between the filling material and the chip, the chip area, and the interface temperature difference of the filling material; the working condition parameter combination includes the chip input power, chip output power, coolant flow rate, maximum coolant flow rate, ambient temperature, and ambient reference temperature; then, use the orthogonal experimental design to generate N groups of experimental data for each type of parameter combination; each type of parameter combination includes the number of experiments, experimental parameters, and the number of levels of each experimental parameter, and the number of levels of each parameter is the same; It should be specifically noted in this embodiment that experimental design methods such as orthogonal experimental design and Latin hypercube sampling are used to generate a series of representative parameter combinations within the determined parameter range; the orthogonal experimental design selects some representative points from the comprehensive experiment according to orthogonality, which is a highly efficient, fast, and economical experimental design method; arrange the experiment according to the L9(3 4 ) orthogonal table (9 experiments, 4 factors, 3 levels for each factor), only 9 experiments are needed. According to the L 15 (3 7 ) orthogonal table, 15 experiments are carried out (15 experiments, 4 factors, 3 levels for each factor). Obviously, the workload is greatly reduced; thus, the orthogonal experimental design has been widely used in the research of many fields; for example, there are 9 experiments for the chip geometric parameter combination, and the experimental parameters include the chip length × width and chip thickness. According to the geometric parameter range, the number of levels of the chip length × width is determined to be 3 (1 mm × 1 mm, 5 mm × 5 mm, 10 mm × 10 mm), and the number of levels of the chip thickness is 3 (0.1 mm, 0.5 mm, 1 mm).

[0018] S3: Through the simulation model construction technology, based on the first text of the heat dissipation structure design and the second text of the heat dissipation structure design obtained in S2, construct a simulation model of the silicon carbide chip heat dissipation structure, including the following steps: S3.1: First, based on the heat dissipation structure designs obtained in S2, namely the first text and the second text of the heat dissipation structure design, obtain the chip geometric parameters and their corresponding parameter ranges, the geometric parameters of heat dissipation and their corresponding parameter ranges, as well as the material property parameters and their corresponding parameter ranges. Then, through simulation model construction techniques (such as SolidWorks software, ANSYS Fluent software, etc.), construct the heat dissipation structure of the silicon carbide chip. Secondly, in combination with the working condition parameters and their corresponding parameter ranges, set the simulation boundary conditions on the basis of the heat dissipation structure of the silicon carbide chip to construct the simulation model of the heat dissipation structure of the silicon carbide chip. It should be specifically noted in this embodiment that SolidWorks is a three-dimensional mechanical design software, widely used in scenarios such as mechanical part design and industrial equipment assembly; ANSYS Fluent is a computational fluid dynamics (CFD) simulation software, used to simulate physical phenomena such as fluid flow, heat transfer, and chemical reactions, and belongs to a part of the ANSYS engineering simulation suite. During the model construction process of this invention application, it can realize complex physical processes such as heat source distribution, heat conduction, convective heat transfer, and radiative heat transfer inside the chip; the geometric model defines the "structure", and the boundary conditions define the "environment", and the two together constitute a complete mathematical description of the simulation problem; based on basic principles such as heat transfer and fluid mechanics, select appropriate numerical simulation methods and software platforms (such as finite element analysis software ANSYS, COMSOL, etc.) to establish the simulation model.

[0019] S3.2: Through big data analysis technology, evaluate the heat dissipation structure of the silicon carbide chip and the simulation boundary condition settings respectively, and obtain the evaluation index GDMP of the heat dissipation structure of the silicon carbide chip and the evaluation index TCRM of the simulation boundary condition settings respectively, where , n represents the number of geometric dimension measurements, n1 and n2 respectively represent the number of types of geometric parameters and material property parameters, D sim,i j1 and D exp,i j1 respectively represent the geometric dimensions in the simulation model and the geometric dimensions in the design for the i-th measurement of the j1-th type of geometric parameter, MA sim,i j2 and MA exp,i j2 respectively represent the material property parameters set in the simulation model and the material property parameters obtained from actual tests for the j2-th type; , n3 represents the number of types of working parameters, B sim,i j3 and B exp,i j3 respectively represent the working condition parameters set in the simulation model and the working condition parameters obtained from actual tests for the j3-th type; S3.3: Through big data analysis technology, if the evaluation index GDMP of the silicon carbide chip heat dissipation structure is less than the corresponding threshold GDMP0, return to S3.1 to reconstruct the silicon carbide chip heat dissipation structure; otherwise, it means the evaluation is passed. If the evaluation index TCRM of the simulation boundary condition setting is less than the corresponding threshold TCRM0 for comparison, return to S3.1 to reset the simulation boundary condition; otherwise, it means the evaluation is passed, and a simulation model of the silicon carbide chip heat dissipation structure is obtained. S4: Input the various parameter combinations obtained in S2 into the simulation model of the silicon carbide chip heat dissipation structure obtained in S3 to construct a multi-dimensional simulation data set, including a chip geometric parameter and heat dissipation performance data set, a heat dissipation structure parameter and heat dissipation efficiency data set, a material property and thermal interface performance data set, and a working condition and dynamic heat dissipation response data set, which includes the following steps: S4.1: Construct a chip geometric parameter and heat dissipation performance data set: First, input the chip geometric parameter combination into the simulation model of the silicon carbide chip heat dissipation structure. Through big data analysis technology, obtain the mapping relationship model between the chip geometric size and the chip thermal resistance R th : , the unit of the thermal resistance Rth is K / W, CT represents the chip thickness, L×W represents the length and width of the chip, a1 represents the thermal conductivity of the chip material (unit: W / (m·K)), a2 represents the convective heat transfer coefficient (unit: W / (m²·K)), and the h distribution is directly output by simulating the flow field and temperature field through software (such as ANSYS Fluent, COMSOL). Then, obtain the chip geometric parameter and heat dissipation performance data set composed of the chip geometric parameter combination and the corresponding chip thermal resistance. It should be specifically noted in this embodiment that the thicker the chip, the greater the thermal resistance; the larger the chip area, the smaller the thermal resistance. The larger the area, the more heat dissipation paths, and the smaller the thermal resistance. The thermal resistance of the chip reflects the difficulty of the chip in conducting the heat generated inside to the external environment. The smaller the thermal resistance, the better the heat dissipation performance of the chip, and the heat can be transferred from the inside of the chip to the outside more smoothly. On the contrary, the larger the thermal resistance, the more difficult it is for the chip to dissipate heat, which may easily lead to an increase in the chip temperature and may affect the performance and reliability of the chip.

[0020] S4.2: Construct a heat dissipation structure parameter and heat dissipation efficiency data set: First, input the heat dissipation structure parameter combination into the simulation model of the silicon carbide chip heat dissipation structure. Through big data analysis technology, obtain the mapping relationship model between the heat dissipation structure parameter and the heat dissipation efficiency η: , η fin represents the fin efficiency, , , a2 represents the convective heat transfer coefficient (unit: W / (m²·K)), tanh() is the hyperbolic tangent function, H fm represents the fin height, S fmDenotes the fin pitch, S opt Denotes the optimal fin pitch, th denotes the fin thickness, λ fin Denotes the thermal conductivity of the fin material (unit: W / (m·K)), N denotes the number of fins; then a dataset of heat dissipation structure parameters and heat dissipation efficiency consisting of combinations of heat dissipation structure parameters and the corresponding heat dissipation efficiency is obtained; It should be specifically noted in this embodiment that the ratio of the height to the pitch of the heat dissipation fins reflects the heat dissipation area density, and the larger the ratio, the higher the efficiency. The shape factor improves the efficiency by increasing the surface area or the turbulence effect (such as wavy fins enhancing convection); the heat dissipation structure fins are a common heat dissipation element in electronic devices, mechanical devices, and other systems that require heat dissipation, and are usually made of high thermal conductivity metal materials (such as aluminum, copper, etc.), and improve the heat dissipation efficiency by increasing the heat dissipation area; in the case of natural convection, S opt ≈5 - 10 mm, and in the case of forced convection, S opt ≈2 - 5 mm.

[0021] S4.3: Construct a dataset of material properties and thermal interface performance: First, input the combination of material property parameters into the simulation model of the silicon carbide chip heat dissipation structure, and through big data analysis technology, obtain the mapping relationship model between the material property parameters and the thermal interface heat flux density q (unit: W / m 2 ): , λ sic and λ fil Denote the thermal conductivity of silicon carbide and the thermal conductivity of the filling material (unit: W / (m·K)), A real Denotes the contact area between the filling material and the chip, A nom Denotes the chip area, obtained by multiplying the length and width of the chip, d denotes the thickness of the filling material (unit: m), ΔT denotes the interface temperature difference of the filling material, and the unit is Kelvin K; then a dataset of material properties and thermal interface performance consisting of combinations of material property parameters and the corresponding thermal interface heat flux density is obtained; S4.4: Construct a dataset of working conditions and dynamic heat dissipation response: First, input the combination of working condition parameters into the simulation model of the silicon carbide chip heat dissipation structure, and through big data analysis technology, obtain the mapping relationship model between the working condition parameters and the heat dissipation response index Hdr: , V flow Denotes the flow rate of the cooling medium, max_V flow Denotes the maximum flow rate of the cooling medium, γ denotes the flow rate influence coefficient, 0.6 ≤ γ ≤ 0.8, β denotes the heat dissipation response influence factor, and β and the exponent γ (such as the 0.7th power of the flow rate) can be fitted through simulation software by CFD simulation or experimental data, 0.5 ≤ β ≤ 0.9, T ab Denotes the ambient temperature, T chipIt represents the temperature of the chip during stable operation, with the temperature unit being Kelvin (K); then a dataset of operating conditions and dynamic heat dissipation responses consisting of combinations of operating condition parameters and corresponding heat dissipation response indices is obtained. In this embodiment, it should be specifically noted that increasing the flow rate enhances convective heat transfer and accelerates temperature stabilization; the ambient temperature drives heat dissipation by affecting the temperature difference, but the effect is weak; the absolute temperature (Kelvin, K), which is a standard practice in thermodynamic calculations; it represents the energy conversion efficiency of the chip. A low energy conversion efficiency indicates a high heat generation rate and a high required heat dissipation efficiency.

[0022] S5: Through big data processing technology, process the multi-dimensional simulation dataset obtained in S4 to obtain a simulation dataset for the heat dissipation structure design after data processing, including processing pass results and processing exception results, and feedback the processing exception results to the design administrator, including the following steps: S5.1: Compare the chip thermal resistance R th with the reference value R th 0 If R th ≤R th 0 , then label the corresponding combination of chip geometric parameters as chip heat dissipation efficient data, and the processing passes. Otherwise, judge whether ΔR th falls within the allowable range. If so, label the corresponding combination of chip geometric parameters as chip heat dissipation good data, and the processing passes. If not, it indicates that the combination of chip geometric parameters has an abnormal processing. S5.2: Compare the heat dissipation efficiency η with the reference value η 0 If η≥η 0 , then label the corresponding combination of heat dissipation structure parameters as heat dissipation structure heat dissipation efficient data, and the processing passes. Otherwise, judge whether Δη falls within the allowable range. If so, label the combination of heat dissipation structure parameters as heat dissipation structure heat dissipation good data, and the processing passes. If not, it indicates that the combination of heat dissipation structure parameters has an abnormal processing. S5.3: Compare the heat flux density q at the thermal interface with the reference value q 0 If q≤q 0 , then label the corresponding combination of material property parameters as material heat dissipation efficient data, and the processing passes. Otherwise, judge whether Δq falls within the allowable range. If so, label the corresponding combination of material property parameters as material heat dissipation good data, and the processing passes. If not, it indicates that the combination of material property parameters has an abnormal processing. S5.4: Compare the heat dissipation response index Hdr with the reference value Hdr 0 If Hdr≥Hdr 0, the corresponding combination of working condition parameters is marked as the data with high heat dissipation efficiency under the working conditions, and the process passes. Otherwise, it is judged whether ΔHdr belongs to the allowable range. If so, the corresponding combination of working condition parameters is marked as the data with good heat dissipation under the working conditions, and the process passes. If not, it indicates that the combination of working condition parameters is abnormally processed; S5.5: Through big data technology, store the processed simulation data set in the database according to the marking results to form a simulation data set for heat dissipation structure design; It should be specifically noted in this embodiment that abnormal data has been removed through data preprocessing technology, such as data with negative heat flux density.

[0023] S6: Through big data analysis technology, evaluate the abnormal processing results to obtain the abnormal evaluation index AEI of the simulation data set for heat dissipation structure design, , nc l represents the number of abnormal processes of the l-th type of parameter combination, Tn l represents the number of experiments of the l-th type of parameter combination. l = 1 represents the combination of chip geometric parameters, l = 2 represents the combination of heat dissipation structure parameters, l = 3 represents the combination of material property parameters, and l = 4 represents the working conditions parameters. If AEI is greater than or equal to the corresponding threshold, it indicates that the construction of the simulation data set for heat dissipation structure design is good. Otherwise, it indicates abnormality, and the abnormal evaluation result is fed back to the design administrator for human-computer interaction, such as unreasonable geometric dimension design, inaccurate material property parameters, etc. Adjust relevant parameters according to the analysis results and re-perform simulation and data acquisition; Secondly: In the attached drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments are involved. Other structures can refer to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other; Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a simulation data set for the heat dissipation structure design of a silicon carbide chip, characterized in that: Including: S1: Through big data technology, collect, store, and update the influencing factors during the heat dissipation process of silicon carbide chips, and receive the data generated in the heat dissipation structure design of silicon carbide chips to construct a heat dissipation structure design database. S2: Based on the heat dissipation structure design database obtained in S1, collect the data required for heat dissipation structure design simulation, and determine various parameter combinations and corresponding experimental data for constructing the simulation data set. S3: Through simulation model construction technology, based on the first text of heat dissipation structure design and the second text of heat dissipation structure design obtained in S2, construct a simulation model of the heat dissipation structure of silicon carbide chips. S4: Input the various parameter combinations obtained in S2 into the simulation model of the heat dissipation structure of silicon carbide chips obtained in S3 to construct a multi-dimensional simulation data set, including chip geometric parameter and heat dissipation performance data set, heat dissipation structure parameter and heat dissipation efficiency data set, material property and thermal interface performance data set, and working condition and dynamic heat dissipation response data set. S5: Through big data processing technology, process the multi-dimensional simulation data set obtained in S4 to obtain a heat dissipation structure design simulation data set after data processing, including processed passing results and processed abnormal results, and feedback the processed abnormal results to the design administrator. S6: Through big data analysis technology, evaluate the processed abnormal results, and feedback according to the abnormal evaluation results to the design administrator for human-computer interaction.

2. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, wherein: The various parameter combinations and corresponding experimental data in S2: include chip geometric parameter combinations, heat dissipation structure parameter combinations, material property parameter combinations, and working condition parameter combinations; among them, the chip geometric parameter combinations include the length, width, and thickness of the chip, the thermal conductivity of the chip material, and the convective heat transfer coefficient; the heat dissipation structure parameter combinations include the convective heat transfer coefficient, fin height, fin pitch, fin thickness, and fin material thermal conductivity; the material property parameter combinations include silicon carbide thermal conductivity and filler material thermal conductivity, the contact area between the filler material and the chip, the chip area, and the interface temperature difference of the filler material; the working condition parameter combinations include chip input power, chip output power, cooling medium flow rate, maximum cooling medium flow rate, ambient temperature, and ambient reference temperature; then use orthogonal experimental design to generate N groups of experimental data for each type of parameter combination; each type of parameter combination includes the number of experiments, experimental parameters, and the number of levels of each experimental parameter, and the number of levels of each parameter is the same.

3. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, characterized in that: In S4, a dataset of chip geometric parameters and heat dissipation performance is constructed: First, the chip geometric parameter combinations are input into the simulation model of the silicon carbide chip heat dissipation structure, and through big data analysis technology, the mapping relationship model between the chip geometric dimensions and the chip thermal resistance R th is obtained: , where CT represents the chip thickness, L×W represents the length and width of the chip, a1 represents the thermal conductivity of the chip material, and a2 represents the convective heat transfer coefficient; then, a dataset of chip geometric parameters and heat dissipation performance composed of chip geometric parameter combinations and the corresponding chip thermal resistances is obtained.

4. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, characterized in that: Construct a dataset of heat dissipation structure parameters and heat dissipation efficiency in S4: First, input the combination of heat dissipation structure parameters into the simulation model of the silicon carbide chip heat dissipation structure. Through big data analysis technology, obtain the mapping relationship model between the heat dissipation structure parameters and the heat dissipation efficiency η: , η fin represents the fin efficiency, , , a2 represents the convective heat transfer coefficient, tanh() is the hyperbolic tangent function, H fm represents the fin height, S fm represents the fin pitch, S opt represents the optimal fin pitch, th represents the fin thickness, λ fin represents the thermal conductivity of the fin material, N represents the number of fins; then obtain a dataset of heat dissipation structure parameters and heat dissipation efficiency composed of the combination of heat dissipation structure parameters and the corresponding heat dissipation efficiency.

5. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, wherein: Constructing the material property and thermal interface performance dataset in S4: First, input the combination of material property parameters into the simulation model of the silicon carbide chip heat dissipation structure. Through big data analysis technology, combined with the thermal conductivity λ of silicon carbide sic and the thermal conductivity λ of the filling material fil , the contact area A between the filling material and the chip real , the chip area A nom , obtained by multiplying the length and width of the chip, the thickness d of the filling material, and ΔT represents the interface temperature difference of the filling material, to obtain the mapping relationship model between the material property parameters and the thermal interface heat flux density q; then obtain the material property and thermal interface performance dataset composed of the combination of material property parameters and the corresponding thermal interface heat flux density.

6. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, characterized in that: In S4, a working condition and dynamic heat dissipation response dataset is constructed: First, the working condition parameter combinations are input into the simulation model of the silicon carbide chip heat dissipation structure. Through big data analysis technology, combined with the cooling medium flow velocity V flow , representing the maximum cooling medium flow velocity max_V flow , the ambient temperature T ab , the temperature T of the chip during stable operation chip and the flow velocity influence coefficient γ, a mapping relationship model between the working condition parameters and the heat dissipation response index Hdr is obtained; then a working condition and dynamic heat dissipation response dataset composed of the working condition parameter combinations and the corresponding heat dissipation response indices is obtained.

7. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, wherein: The heat dissipation structure design simulation data set after data processing obtained in S5: includes: S5.1: Compare the chip thermal resistance R th with the reference value R th 0 . If R th ≤R th 0 , label the corresponding chip geometric parameter combination as efficient chip heat dissipation data and the process passes. Otherwise, determine whether ΔR th falls within the allowable range. If so, label the corresponding chip geometric parameter combination as good chip heat dissipation data and the process passes. If not, it indicates that the chip geometric parameter combination is abnormally processed; S5.2: Compare the heat dissipation efficiency η with the reference value η 0 and if η ≥ η 0 , label the corresponding combination of heat dissipation structure parameters as high-efficiency heat dissipation data of the heat dissipation structure, and the process passes. Otherwise, determine whether Δη belongs to the allowable range. If so, label the combination of heat dissipation structure parameters as good heat dissipation data of the heat dissipation structure, and the process passes. If not, it indicates that the processing of the combination of heat dissipation structure parameters is abnormal.

8. The method for constructing a simulation data set based on the heat dissipation structure design of a silicon carbide chip according to claim 1, characterized in that: The heat dissipation structure design simulation data set after data processing obtained in S5: also includes: S5.3: Compare the thermal interface heat flux density q with the reference value q 0 and if q ≤ q 0 , then label the corresponding combination of material property parameters as highly efficient heat dissipation data of the material, and the process passes. Otherwise, determine whether Δq falls within the allowable range. If so, label the corresponding combination of material property parameters as good heat dissipation data of the material, and the process passes. If not, it indicates an abnormality in the processing of the combination of material property parameters; S5.4: Compare the heat dissipation response index Hdr with the reference value Hdr 0 If Hdr ≥ Hdr 0 , label the corresponding combination of working condition parameters as high-efficiency heat dissipation data for working conditions, and the process passes. Otherwise, determine whether ΔHdr falls within the allowable range. If so, label the corresponding combination of working condition parameters as good heat dissipation data for working conditions, and the process passes. If not, it indicates an abnormality in the processing of the combination of working condition parameters; S5.5: Through big data technology, store the passed simulation data set in the database according to the annotation results to form a heat dissipation structure design simulation data set.

Images