Interface thermal resistance test design method, test device and evaluation method

Abstract

The invention discloses an interface thermal resistance test design method, a test device and an evaluation method, and belongs to the technical field of electronic equipment thermal management, and the method comprises the steps: S1, determining factors affecting the interface thermal resistance according to the installation condition of a heat conduction insulation pad in an electronic module; s2, determining the level of each factor according to engineering installation and use conditions; s3, selecting an orthogonal test table according to the determined factors and levels, and determining a test scheme and a sample piece; s4, testing the interface thermal resistance of the sample piece; and S5, carrying out analysis according to a test result, determining main factors influencing the interface thermal resistance, fitting a relationship between each influence factor and the interface thermal resistance by using a neural network, and determining a respective variable parameter value, so that rapid evaluation of the contact thermal resistance under the corresponding installation condition can be realized. According to the method, the defect of low matching degree between a conventional test method and actual installation is overcome, the test efficiency is improved, the interface thermal resistance can be accurately mastered, the heat dissipation evaluation accuracy of the electronic device is improved, and refined thermal management of the electronic module is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal management of electronic equipment, and more specifically, to an interface thermal resistance test design method, a test device and an evaluation method. Background Art

[0002] With the development of microelectronics technology, the power density of electronic devices has increased significantly, and the problem of device heat dissipation has gradually become a difficult problem that needs to be solved urgently for electronic equipment. Interface materials represented by compressible thermally conductive insulating pads are often used inside electronic modules. They are attached between the device and the heat sink surface to transfer the heat of the device to the heat sink surface to achieve device heat dissipation. The interface contact thermal resistance value between the device and the heat sink surface is an important parameter for evaluating the heat dissipation of electronic devices. At present, the contact thermal resistance evaluation of thermally conductive insulating pads is mostly obtained through the pressure-thermal resistance curve. As described in patent application CN202211167644.0, this evaluation method has a low degree of matching with the actual use scenario. In actual use, the pressure distribution caused by different installation structures is uneven, and the compression pressure is difficult to accurately obtain, resulting in inaccurate thermal resistance evaluation. Therefore, it is urgent to carry out compressible thermally conductive insulating pad interface thermal resistance testing and evaluation solutions that match the actual installation. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an interface thermal resistance test design method, a test device and an evaluation method, which solves the shortcoming that the conventional test method has a low matching degree with the actual installation, improves the test efficiency, can accurately grasp the interface thermal resistance, improves the accuracy of the heat dissipation evaluation of electronic devices, and realizes refined thermal management of electronic modules.

[0004] The object of the present invention is achieved through the following solutions:

[0005] An interface thermal resistance test design method is provided, which is based on the orthogonal test design of the interface thermal resistance test of the compressible thermal conductive insulation pad, and specifically includes:

[0006] Step S1, determining factors affecting interface thermal resistance according to the installation condition of the thermally conductive insulating pad in the electronic module;

[0007] Step S2, determining the level of each factor according to the project installation and use conditions;

[0008] Step S3, selecting an orthogonal test table and determining a test plan and samples according to the factors and levels determined in steps S1 and S2;

[0009] Step S4, conducting a sample interface thermal resistance test;

[0010] Step S5, analyzing the test results to determine the main factors affecting the interface thermal resistance.

[0011] Further, in step S1, the factors include screw spacing P, screw specification D, interface spacing l and roughness Ra.

[0012] An interface thermal resistance testing device comprises: a heating plate, a cold plate, a compressible thermal pad, an insulating gasket and mounting screws; the compressible thermal pad is installed between the heating plate and the cold plate, and the thickness of the insulating gasket is used to control the compressed size of the compressible thermal pad, thereby controlling the compression amount; the heating plate and the cold plate are locked by the mounting screws to provide locking pressure; during the test, a heat source is applied to the heating plate, the temperature at the contact point between the heating plate, the cold plate and the compressible thermal pad is measured, and the interface thermal resistance is calculated; and the device is used to complete the interface thermal resistance test in the interface thermal resistance test design method as described above.

[0013] Furthermore, the thermal insulation gasket is made of low thermal conductivity material to reduce lateral heat loss.

[0014] A method for evaluating interface thermal resistance adopts a BP neural network to train the test results of the above test device to obtain the corresponding relationship between the influencing factors and the interface thermal resistance, and uses the trained neural network to evaluate the interface thermal resistance value under any combination of influencing factors.

[0015] Furthermore, the BP neural network is used to train the test results in claim 3 to obtain the corresponding relationship between the influencing factors and the interface thermal resistance, and the trained neural network is used to evaluate the interface thermal resistance value under any combination of influencing factors, specifically including the following sub-steps:

[0016] Step 1: Obtain the interface thermal resistance test results and obtain the basic data set for neural network training;

[0017] Step 2, determine the neural network input layer as the factors affecting the interface thermal resistance, and the number of neurons in the input layer is the number of types of influencing factors;

[0018] Step 3, determine that the output layer of the neural network is the interface thermal resistance value obtained by the test, and the number of neurons in the output layer is 1;

[0019] Step 4: Conduct independence verification of the number of hidden layer neurons, select 1-n neurons as the verification range, and obtain the root mean square error MSE and regression coefficient Reg of the training results when the number of neurons is k, where k is an integer from 1 to n;

[0020] Step 5: Based on the minimum root mean square error (MSE) and the regression coefficient (Reg) closest to 1, the number of hidden layer neurons with the highest training accuracy is selected;

[0021] Step 6, performing neural network training with the training parameters determined in Step 2, Step 3, and Step 5, and obtaining a neural network net of each factor combination and interface thermal resistance;

[0022] Step 7: Input the combination value of the factors to be evaluated for thermal resistance;

[0023] Step eight, using the neural network trained in step six, fit the input of step seven and output the interface thermal resistance evaluation value.

[0024] The beneficial effects of the present invention include:

[0025] The present invention fully considers the actual installation scenario of the thermal pad by changing factors such as the screw spacing P, the screw specification D, the interface spacing l, and the roughness Ra, and solves the problem that the conventional test method has a low matching degree with the actual installation. The influence of various factors on the interface thermal resistance is obtained by experiment; the test matrix is ​​simplified by the orthogonal test method to obtain the test array table, which reduces the number of samples while fully studying the influence of various parameters and improves the test efficiency; with the test results as the basic database, a neural network is used to fit the relationship between various influencing factors and the interface thermal resistance, and the parameter values ​​of each variable are determined to achieve rapid evaluation of the contact thermal resistance under the corresponding installation conditions.

[0026] In an embodiment of the present invention, a method for designing an interface thermal resistance test for a compressible thermally conductive insulating pad based on an orthogonal test is proposed to measure the interface thermal resistance values ​​under different levels of different factors. The orthogonal test is used to reduce the number of test samples, effectively reduce the test cost, and improve the test efficiency. The factors that significantly affect the interface thermal resistance are obtained through analysis of the orthogonal test method.

[0027] In the embodiment of the present invention, an interface thermal resistance test device based on the actual installation scenario of the thermal conductive insulation pad is proposed. The clamping force is provided by the locking screw, which is different from the traditional uniform loading pressure method. By changing the values ​​of various factors, the matching degree between the measurement method and the actual use scenario is improved, thereby improving the accuracy of the interface thermal resistance measurement, providing basic data for accurate thermal simulation evaluation, and finally realizing refined thermal management. The compression amount is controlled by the insulating gasket, reducing the complexity of the test device, while effectively isolating the heat from the side to improve the measurement accuracy.

[0028] In an embodiment of the present invention, a method for quickly evaluating the interface thermal resistance of a compressible thermal conductive pad based on a neural network is proposed. Based on the measurement data, the interface thermal resistance evaluation of multiple factors under different parameter combinations is realized through neural network training, thereby improving the efficiency and accuracy of thermal resistance evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0030] Figure 1 This is a flow chart of a method for testing thermal resistance of an interface according to an embodiment of the present invention;

[0031] Figure 2a is the first schematic diagram of the test sample;

[0032] Figure 2b is the second schematic diagram of the test sample;

[0033] Figure 3 It is a flow chart for thermal resistance evaluation;

[0034] In the figure, 1-heating plate; 2-cold plate; 3-compressible thermal pad; 4-insulating gasket; 5-mounting screws. DETAILED DESCRIPTION

[0035] All features disclosed in all embodiments in this specification, or steps in all methods or processes implicitly disclosed, except for mutually exclusive features and / or steps, can be combined and / or expanded or replaced in any manner.

[0036] In view of the problems in the background, in the conception of the present invention, a thermal resistance test design, test and evaluation scheme for a compressible thermally conductive insulating pad for electronic module heat dissipation is proposed. By changing factors such as screw spacing P, screw specification D, interface spacing l, roughness Ra, etc., the actual installation scenario of the thermal conductive pad can be fully considered to solve the problem of low matching between conventional test methods and actual installation. The influence of various factors on the interface thermal resistance is obtained by the experiment; the test matrix is ​​simplified by the orthogonal test method to obtain the test array table, which reduces the number of samples while fully studying the influence of various parameters and improves the test efficiency; based on the test results as the basic database, a neural network is used to fit the relationship between various influencing factors and interface thermal resistance, and the parameter values ​​of each variable are determined to achieve rapid evaluation of contact thermal resistance under corresponding installation conditions.

[0037] In a specific embodiment, the screw spacing P is 30mm, 50mm, 70mm, and 90mm, the screw specifications D are M2, M2.5, M3, and M4, the interface spacing l is 0.5mm, 0.75mm, 1mm, and 1.25mm, and the surface roughness Ra is 0.8, 1.6, 3.2, and 6.3. Considering three materials, the interface thermal resistance test is carried out. The total sample test volume is 768 groups, and the sample volume is reduced to 16*3=48 groups by orthogonal test method. The measured interface thermal resistance is used as training data, and the BP neural network is used for training. The number of neurons in the input layer is 4, the number of neurons in the output layer is 1, and the number of neurons in the hidden layer is 9. The interface thermal resistance of a thermal pad is evaluated under the installation conditions of screw M2, spacing of 40mm, interface spacing of 0.6mm, and aluminum plate roughness of Ra1.6. That is, (2, 40, 0.6, 1.6) is used as the input value, and the trained neural network is used for fitting, and the interface thermal resistance is 4.41W / cm2 K.

[0038] It should be noted that within the scope of protection defined in the claims of the present invention, the following embodiments can be combined and / or expanded or replaced in any logical way from the above specific implementation methods, such as disclosed technical principles, disclosed technical features or implicitly disclosed technical features.

[0039] Example 1

[0040] like Figure 1 As shown, an interface thermal resistance test design method is provided, and the interface thermal resistance test design of a compressible thermal conductive insulation pad based on an orthogonal test specifically includes:

[0041] Step S1, determining factors affecting interface thermal resistance according to the installation condition of the thermally conductive insulating pad in the electronic module;

[0042] Step S2, determining the level of each factor according to the project installation and use conditions;

[0043] Step S3, selecting an orthogonal test table and determining a test plan and samples according to the factors and levels determined in steps S1 and S2;

[0044] Step S4, conducting a sample interface thermal resistance test;

[0045] Step S5, analyzing the test results to determine the main factors affecting the interface thermal resistance.

[0046] Example 2

[0047] On the basis of Example 1, in step S1, the factors include screw spacing P, screw specification D, interface spacing l and roughness Ra.

[0048] Example 3

[0049] like Figure 2a and Figure 2b As shown, an interface thermal resistance testing device comprises: a heating plate 1, a cold plate 2, a compressible thermal pad 3, an insulating gasket 4 and a mounting screw 5; the compressible thermal pad 3 is installed between the heating plate 1 and the cold plate 2, and the thickness of the insulating gasket 4 is used to control the compressed size of the compressible thermal pad 3, thereby controlling the compression amount; the heating plate 1 and the cold plate 2 are locked by the mounting screws 5 to provide locking pressure; during the test, a heat source is applied to the heating plate 1, and the temperature at the contact point between the heating plate 1, the cold plate 2 and the compressible thermal pad 3 is measured to calculate the interface thermal resistance; and the device is used to complete the interface thermal resistance test in the interface thermal resistance test design method described in claim 1.

[0050] Example 4

[0051] Based on Example 3, the thermal insulation gasket 4 is made of low thermal conductivity material to reduce lateral heat loss.

[0052] Example 5

[0053] A method for evaluating interface thermal resistance adopts a BP neural network to train the test results in Example 3 to obtain the corresponding relationship between influencing factors and interface thermal resistance, and uses the trained neural network to evaluate the interface thermal resistance value under any combination of influencing factors.

[0054] Example 6

[0055] On the basis of Example 5, the BP neural network is used, such as Figure 3 As shown, the test results in Example 3 are trained to obtain the corresponding relationship between the influencing factors and the interface thermal resistance, and the trained neural network is used to evaluate the interface thermal resistance value under any combination of influencing factors, which specifically includes the following sub-steps:

[0056] Step 1: Obtain the interface thermal resistance test results and obtain the basic data set for neural network training;

[0057] Step 2, determine the neural network input layer as the factors affecting the interface thermal resistance, and the number of neurons in the input layer is the number of types of influencing factors;

[0058] Step 3, determine that the output layer of the neural network is the interface thermal resistance value obtained by the test, and the number of neurons in the output layer is 1;

[0059] Step 4: Conduct independence verification of the number of hidden layer neurons, select 1-n neurons as the verification range, and obtain the root mean square error MSE and regression coefficient Reg of the training results when the number of neurons is k, where k is an integer from 1 to n;

[0060] Step 5: Based on the minimum root mean square error (MSE) and the regression coefficient (Reg) closest to 1, the number of hidden layer neurons with the highest training accuracy is selected;

[0061] Step 6, performing neural network training with the training parameters determined in Step 2, Step 3, and Step 5, and obtaining a neural network net of each factor combination and interface thermal resistance;

[0062] Step 7: Input the combination value of the factors to be evaluated for thermal resistance;

[0063] Step eight, using the neural network trained in step six, fit the input of step seven and output the interface thermal resistance evaluation value.

[0064] The units involved in the embodiments of the present invention may be implemented by software or hardware, and the units described may also be arranged in a processor. The names of these units do not, in some cases, limit the units themselves.

[0065] According to one aspect of an embodiment of the present invention, a computer program product or a computer program is provided, the computer program product or the computer program includes a computer instruction, and the computer instruction is stored in a computer-readable storage medium. A processor of a computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the method provided in the above various optional implementations.

[0066] As another aspect, an embodiment of the present invention further provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment.

Claims

1. An interface thermal resistance test design method, characterized in that: The design of the thermal resistance test of the compressible thermal conductive insulation pad interface based on orthogonal test includes: Step S1, determining factors affecting interface thermal resistance according to the installation condition of the thermally conductive insulating pad in the electronic module; Step S2, determining the level of each factor according to the project installation and use conditions; Step S3, selecting an orthogonal test table and determining a test plan and samples according to the factors and levels determined in steps S1 and S2; Step S4, conducting a sample interface thermal resistance test; Step S5, analyzing the test results to determine the main factors affecting the interface thermal resistance.

2. The interface thermal resistance test design method according to claim 1, characterized in that: In step S1, the factors include screw pitch P, screw specification D, interface pitch l and roughness Ra.

3. An interface thermal resistance testing device, characterized in that: include: A heating plate (1), a cold plate (2), a compressible thermal pad (3), an insulating gasket (4) and mounting screws (5); the compressible thermal pad (3) is mounted between the heating plate (1) and the cold plate (2), and the thickness of the insulating gasket (4) is used to control the compressed size of the compressible thermal pad (3), thereby controlling the compression amount; the heating plate (1) and the cold plate (2) are locked by the mounting screws (5) to provide locking pressure; during testing, a heat source is applied to the heating plate (1), the temperature at the contact point between the heating plate (1), the cold plate (2) and the compressible thermal pad (3) is measured, and the interface thermal resistance is calculated; and the device is used to complete the interface thermal resistance test in the interface thermal resistance test design method described in claim 1.

4. The interface thermal resistance testing device according to claim 3, characterized in that: The thermal insulation gasket (4) is made of low thermal conductivity material and is used to reduce lateral heat loss.

5. A method for evaluating interface thermal resistance, characterized in that: The BP neural network is used to train the test results in claim 3 to obtain the corresponding relationship between the influencing factors and the interface thermal resistance, and the trained neural network is used to evaluate the interface thermal resistance value under any combination of influencing factors.

6. The interface thermal resistance evaluation method according to claim 5, characterized in that: The method of using a BP neural network to train the test results in claim 3 to obtain the corresponding relationship between the influencing factors and the interface thermal resistance, and using the trained neural network to evaluate the interface thermal resistance value under any combination of influencing factors, specifically includes the following sub-steps: Step 1: Obtain the interface thermal resistance test results and obtain the basic data set for neural network training; Step 2, determine the neural network input layer as the factors affecting the interface thermal resistance, and the number of neurons in the input layer is the number of types of influencing factors; Step 3, determine that the output layer of the neural network is the interface thermal resistance value obtained by the test, and the number of neurons in the output layer is 1; Step 4: Conduct independence verification of the number of hidden layer neurons, select 1-n neurons as the verification range, and obtain the root mean square error MSE and regression coefficient Reg of the training results when the number of neurons is k, where k is an integer from 1 to n; Step 5: Based on the minimum root mean square error (MSE) and the regression coefficient (Reg) closest to 1, the number of hidden layer neurons with the highest training accuracy is selected; Step 6, performing neural network training with the training parameters determined in Step 2, Step 3, and Step 5, and obtaining a neural network net of each factor combination and interface thermal resistance; Step 7: Input the combination value of the factors to be evaluated for thermal resistance; Step eight, using the neural network trained in step six, fit the input of step seven and output the interface thermal resistance evaluation value.

Images