A high heat rate SiC MOSFET power device thermal analysis method based on finite element method

Abstract

The application discloses a high-heat-consumption SiC MOSFET power device thermal analysis method based on a finite element method, which comprises the following steps: step 1, loading material information; step 2, three-dimensional modeling, modeling according to the three-dimensional size and arrangement position of each component to obtain a three-dimensional model; step 3, model correction, performing arc processing and simplification processing on the edges and corners of the three-dimensional model; dividing a calculation grid based on the simplified model; step 4, setting, setting the material properties of each part of the simplified model; setting the boundary conditions of the simplified model; step 5, obtaining a steady-state temperature, selecting a solid heat transfer physical field model, and using a finite element numerical method of a heat conduction equation to obtain the steady-state temperature of the high-heat-consumption SiC MOSFET power device to be analyzed. The application can more efficiently calculate the working steady-state temperature of the high-heat-consumption SiC MOSFET power device, and the simplified SiC MOSFET model used can improve the calculation speed while meeting the engineering error requirement.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of semiconductor technology, and particularly relates to a high-thermal-loss SiC MOSFET power device thermal analysis method based on a finite element method. BACKGROUND

[0002] In modern power electronic systems, SiC MOSFETs are becoming the first choice for power devices. Compared with traditional Si-based devices, SiC MOSFETs have excellent properties such as low switching loss, high switching frequency, and high operating temperature, so SiC MOSFETs have been widely used in emerging and traditional fields such as electric vehicles, 5G base stations, and power systems. However, as power devices are developing towards high power and small size, they will inevitably face increasingly serious thermodynamic problems. The operating temperature has a huge impact on the performance and reliability of electronic devices, and existing research shows that high temperature-induced thermal stress is one of the main causes of power semiconductor device failure, so the research on thermal analysis simulation technology of power devices is of great significance.

[0003] The purpose of thermal analysis is to obtain the thermal characteristics of SiC MOSFET and other power semiconductors during operation, and to serve as a reference for optimizing the design of heat dissipation systems and reliability evaluation. When performing thermal analysis of SiC MOSFET, numerical simulation methods such as finite element analysis (FEA) can be used to establish a three-dimensional thermal model of the MOSFET, taking into account its internal structure, material properties, and operating conditions, and to calculate its temperature distribution, thermal resistance, and heat conduction parameters.

[0004] The solution method for thermal analysis using the finite element method is to discretize the entity by establishing a grid, and for each grid element, to establish energy conservation and heat conduction, heat convection equations, and to solve the balance equations of the entire entity to obtain the temperature value of each grid as the result of the simulation. The finite element thermal analysis method has high reliability and short calculation time. Based on finite element simulation, the device structure design can be optimized, the heat dissipation system can be improved, and the working efficiency and reliability of the device can be improved.

[0005] Current researches are usually focused on the optimization of single SiC MOSFET packaging technology, or the research objects are mostly low-thermal-loss devices with sufficient heat dissipation space. However, as SiC MOSFETs are increasingly integrated into various fields, the challenges of high power and poor heat dissipation conditions become more and more urgent to overcome. Therefore, there is an urgent need for a thermal analysis method for high-thermal-loss power devices based on the finite element method. SUMMARY

[0006] To adapt to the above-mentioned needs, the application aims to provide a thermal analysis method for high heat power devices based on the finite element method, a simplified SiC MOSFET model is constructed to fit the actual heat source, and the thin layer structure is optimized when the overall model is established, so that the final steady-state result is obtained more quickly when the finite element simulation is performed, and the accuracy is also ensured.

[0007] To achieve the above-mentioned purpose, the technical scheme adopted by the application is as follows:

[0008] A high heat rate SiC MOSFET power device thermal analysis method based on the finite element method, comprising the following steps:

[0009] Step 1, load material information, the material includes the material contained in each component of the high heat rate SiC MOSFET power device to be analyzed;

[0010] Step 2, three-dimensional modeling, modeling according to the three-dimensional size and arrangement position of each component, obtaining a three-dimensional model;

[0011] Step 3, model correction, rounding and simplifying the edges and corners of the three-dimensional model, and when there is a segmented structure, the segmented structure is merged into a whole or part to obtain a simplified model; based on the simplified model, the calculation grid is divided;

[0012] Step 4, setting, setting the material properties of each part of the simplified model; setting the boundary conditions of the simplified model;

[0013] Step 5, obtaining steady-state temperature, selecting a solid heat transfer physical field model, and using the finite element numerical method of heat conduction equation to obtain the steady-state temperature of the high heat rate SiC MOSFET power device to be analyzed.

[0014] Preferably, step 2 comprises the following steps:

[0015] Step 2.1, ignore the internal structure of each component in three-dimensional modeling, simplify the theoretical thermal resistance of the model, and calculate according to the 45° heat conduction model;

[0016] Step 2.2, when three-dimensional modeling, ignore the part with thickness much smaller than other devices, convert it into a face boundary condition, and set the corresponding physical properties.

[0017] Preferably, the construction method of the simplified model is as follows:

[0018] Two cuboids made of SiC and Cu are set up, with dimensions equivalent to the actual product. A square at the center of the interface is selected as the heat source. Pins are added to the model, with the number and material of the pins being the same as those of the high heat dissipation SiC MOSFET power device being analyzed.

[0019] More preferably, the side length of the square is 4mm.

[0020] Preferably, the theoretical thermal resistance calculation method for the simplified model is as follows:

[0021]

[0022] Where: L i It is the thickness of the i-th layer of material; a i b i These represent the length and width of the lower surface of the i-th layer of material, respectively; K i Let R be the thermal conductivity of the i-th layer material, then the theoretical junction-to-case thermal resistance R of the SiC MOSFET power chip is... jc for:

[0023] R jc =Σ i R thi .

[0024] Furthermore, the theoretical thermal resistance calculation method for the simplified model also includes:

[0025] Compare and select SiC MOSFET datasheets for R jc The theoretical R of the above model jc This ensures that the error is within a reasonable range.

[0026] Preferably, the error is within a reasonable range, meaning the error is less than 20%.

[0027] Preferably, step 4 includes the following steps:

[0028] Step 4.1: Set the heat convection coefficient of the exposed contact surface;

[0029] Step 4.2: Simplify the model into a thin layer with surface boundary conditions, set the corresponding material properties, and set the thermal convection coefficient of the thin layer;

[0030] Step 4.3: Simplify the radiator model by replacing it with an equivalent method that increases the thermal convection coefficient of the area where the radiator is located.

[0031] A further preferred embodiment is characterized in that,

[0032] In step 4.1, the thermal convection coefficient of the exposed contact surface is set to 5-20 (W / m²).2 K);

[0033] In step 4.2, the heat convection coefficient of the thin layer is set to 0;

[0034] Preferably, the heat conduction equation is as follows:

[0035]

[0036] Where, p is the density, C p is the specific heat capacity, u is the velocity vector, is the temperature gradient, is the divergence of the heat flux density vector q, k is the thermal conductivity, Q and Q ted are the internal and external heat sources, respectively.

[0037] The beneficial effects of the present application are as follows:

[0038] The present application can more efficiently calculate the working steady-state temperature of a SiC MOSFET power device with high heat consumption rate, and the simplified SiC MOSFET model used can improve the calculation speed while meeting the engineering error requirements. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 is a flowchart of the high heat consumption rate SiC MOSFET power device thermal analysis method disclosed in the present application based on the finite element method. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application will be further described in detail below with reference to the accompanying drawings.

[0041] Example 1

[0042] The present embodiment discloses a high heat consumption rate SiC MOSFET power device thermal analysis method based on the finite element method, as shown in Figure 1 , specifically comprising the following steps:

[0043] Step 1: Set the material: according to the predetermined working temperature, the physical parameters of the material should be within a reasonable range.

[0044] Step 2: Establish a three-dimensional model of the power device, according to the design of the device, determine the three-dimensional size and arrangement position of each component, and model it in the finite element software, specifically:

[0045] Step 2.1: Establish a three-dimensional model of each component of the power device in turn, and the main dimension of each three-dimensional model needs to be equal to the actual product.

[0046] Step 2.2: For high heat rate SiC MOSFET power devices, SiC MOSFET is the main heat source, and the internal structure is complex, so its three-dimensional model needs to be simplified, ignoring its internal structure. Set two cuboids of SiC and Cu materials respectively, and the dimensions of the two cuboids are temporarily set to be equal to the actual product, and select a square of 4mm in length in the middle of the interface as the heat source; the model needs to add copper pins, the number and material are equal to the selected product.

[0047] Step 2.3: The theoretical thermal resistance of the simplified model can be calculated according to the 45° heat conduction model, and the error with the given nominal value of the selected device should be within the engineering allowable range. If the theoretical thermal resistance of the simplified model temporarily set at present has met the engineering error condition, it is not necessary to modify, if not, the dimensions of the two cuboids of SiC and Cu need to be modified, and the modification is in the order of thickness, then length and width, and the modification value of length and width should not be too large. The calculation method of the theoretical thermal resistance is as follows:

[0048]

[0049] Where, L i is the thickness of the i-th layer of material; a i , b i are the length and width of the lower surface of the i-th layer of material, respectively; K i is the thermal conductivity of the i-th layer of material, then the theoretical thermal resistance R jc from the junction to the shell of the power chip of SiC MOSFET is:

[0050] R jc = Σ i R thi

[0051] Compare the R jc in the SiC MOSFET data manual with the theoretical R jc of the above model to ensure that the error is within a reasonable range.

[0052] Step 2.4: In the 3D model, the part with a thickness much smaller than other devices is ignored, which is converted into a surface boundary condition, and the corresponding physical properties are set.

[0053] Step 3: Modify the model, round the corners of the three-dimensional model, and for complex segmented structures, they can be partially merged into a whole; based on the simplified model, divide the calculation grid, and the divided grid should ensure reasonable density and size change rate, as follows:

[0054] Step 3.1: After the three-dimensional model is established, the sharp structure and the small size of the bending structure of the model need to be rounded to reduce the occurrence of corners, which is convenient for the division of the grid and the final solution. For example, the pin of the SiC MOSFET model is usually bent and then welded when it contacts the PCB board, which causes the pin to have a repeated bending shape during modeling. If the rounding process is not performed, the grid density at this location will be large, the grid change rate will be too fast, and the grid quality will be poor and cannot be solved.

[0055] Step 3.2: The simplified model is divided into calculation grids, and the quality of the grid is related to the calculation speed and result. The unit shape, grid density, and grid type should be considered, as follows:

[0056] Grid type: The correct type of grid is selected according to the model. In the high heat rate SiC MOSFET power device thermal analysis method based on the finite element method proposed in this case, although part of the surface boundary condition is simplified separately, the grid for this part of the surface boundary condition is not set separately, but is considered as part of the 3D grid.

[0057] Unit shape: The divided grid unit should avoid irregular shapes. In this embodiment, regular 3D grids such as regular hexahedrons are used to reduce shape errors and improve calculation accuracy. The aspect ratio of the grid unit should be close to 1 to avoid precision loss caused by a large aspect ratio. For some complex structures, such as the pin part of the SiC MOSFET, the unit shape can not be strictly required.

[0058] Grid density: The grid density should be adjusted according to the characteristics of different regions of the model. In this embodiment, in the region where the heat flux gradient changes greatly, such as the heat flux concentration area near the heat source or the boundary layer, the grid density should be increased to obtain a more detailed temperature distribution cloud chart; in the region where the change is gentle, a grid with a larger change rate can be used to reduce the resources occupied during calculation and improve the calculation speed.

[0059] Grid refinement: In this embodiment, the adaptive grid refinement method is selected, which determines the area that needs to be refined through error estimation.

[0060] The grid division of this embodiment needs to consider the accuracy of the result, the convergence of the calculation result, and the calculation resources to achieve the best effect.

[0061] Step 4: Set the material properties of each part of the modified model; set the boundary conditions of the model, which should conform to the actual physical conditions; equivalent the thin layer structure into a surface boundary condition and set the corresponding attribute parameters; equivalent the heat sink model into the additional convective coefficient of the heat sink contact surface, and determine the range of the convective coefficient according to the selected heat sink type, as follows:

[0062] Step 4.1: Set the heat convection coefficient of the model contact surface exposed to air to 5-20 (W / m 2 K), select the appropriate coefficient in the range according to the corresponding environmental conditions;

[0063] Step 4.2: The thin layer simplified into a surface boundary condition during modeling, such as the solder layer, also needs to set the corresponding material properties, but since it is a contact surface, its heat convection coefficient needs to be set to 0;

[0064] Step 4.3: Simplify the heat sink model, and replace it with an equivalent increase in the heat convection coefficient of the area where the heat sink is located. This embodiment provides two equivalent heat convection coefficient values for the heat sink model, the equivalent heat convection coefficient for forced air cooling is 100 (W / m 2 K), and the equivalent heat convection coefficient for external heat sink combined with air cooling is 550 (W / m 2 K).

[0065] Step 5: Select the solid heat transfer physical field model in the finite element software. The three main ways of heat transfer are heat conduction, heat radiation, and heat convection. The main form of heat transfer between solids and gases with temperature difference is heat convection, which is the heat exchange between solids and the environment air. This part has given the corresponding parameter settings in the interface above, while the main heat transfer between each solid grid element is still heat conduction, and the heat convection and heat radiation can be ignored. Therefore, only the finite element numerical method of the steady-state heat conduction equation can be used to solve the steady-state temperature of the power device, as follows:

[0066] The heat conduction equation for solving the steady-state temperature of the model is:

[0067]

[0068] Where, p is the density, C p is the specific heat capacity, u is the velocity vector, is the temperature gradient, is the divergence of the heat flux vector q, k is the thermal conductivity, Q and Q ted are the internal and external heat sources, respectively.

[0069] Finally, by establishing energy conservation and heat conduction, heat convection equations for each grid element respectively, and solving the balance equation of the whole entity, the temperature value of each grid is obtained as the simulation result.

[0070] Embodiment 2

[0071] In order to better understand and implement the present application, on the basis of embodiment 1, this embodiment discloses a thermal analysis example of a SiC MOSFET power device with high heat dissipation rate, as follows:

[0072] Firstly, the information of each component used by the power device is determined, including material and main dimension, as shown in Table 1:

[0073] Table 1: Model information

[0074]

[0075] The material and main dimension of the main device are shown in Table 1. For the solder layer, since its thickness is small, it is set as a surface boundary condition, which is set to a small value, and can be set more specifically through the surface condition interface.

[0076] The physical parameters of the material are loaded into the finite element simulation software, and then a three-dimensional model is established through the modeling system in the software. For the model of SiC MOSFET, it is established in the form of a simplified model, i.e. superimposed by two cuboids of SiC and Cu, and the heat source is set as a square with a side length of 4mm at the interface.

[0077] Subsequently, the model is processed with a circular arc, and the grid is divided. The grid adopted is a regular hexagonal grid, and a regular quadrilateral grid is adopted at the bent pin. The change speed of the grid is 1.10.

[0078] The equivalent heat convection coefficient of different heat sink models is shown in Table 2:

[0079] Table 2: Equivalent heat convection coefficient of different heat sink models

[0080]

[0081] In the finite element software interface, the solid heat transfer and heat conduction steady-state equation is selected according to embodiment 1. The error between the final simulation result and the test result is within the acceptable range.

[0082] Of course, the present application can have other various embodiments. Those skilled in the art can make various corresponding changes and modifications to the present application without departing from the spirit and essence of the present application, but these corresponding changes and modifications should all belong to the protection scope of the claims attached to the present application.

Claims

1. A finite element method-based high-heat-rate SiC MOSFET power device thermal analysis method, characterized by, The method comprises the following steps: Step 1, loading material information, the material including materials contained in each component of a high heat consumption SiC MOSFET power device to be analyzed; Step 2, three-dimensional modeling, modeling according to the three-dimensional size and arrangement position of each component, obtaining a three-dimensional model; Step 3, model correction, rounding and simplifying the edges and corners of the three-dimensional model, and when there is a segmented structure, merging the segmented structure as a whole or part to obtain a simplified model; based on the simplified model, dividing the calculation grid; Step 4, setting, setting the material properties of each part of the simplified model; setting the boundary conditions of the simplified model; Step 5, obtaining steady-state temperature, selecting a solid heat transfer physical field model, and using the finite element numerical method of heat conduction equation to obtain the steady-state temperature of the high heat consumption SiC MOSFET power device to be analyzed; The construction method of the simplified model is as follows: Two cuboids with materials of SiC and Cu are set, the size of the two cuboids is equal to the actual product, and a square in the middle of the interface is selected as a heat source; add pins to the model, the number and material of the pins are equal to the number and material of the pins of the high heat consumption SiC MOSFET power device to be analyzed; The theoretical thermal resistance calculation method of the simplified model is as follows: ， wherein: is the thickness of the i-th layer of material; , are the length and width of the lower surface of the i-th layer of material, respectively; is the thermal conductivity of the i-th layer of material, then the theoretical thermal resistance from junction to case of the power chip of the SiC MOSFET is . 。 2. The thermal analysis method of a high-heat SiC MOSFET power device based on the finite element method according to claim 1, wherein, Step 2 comprises the following steps: Step 2.1, in three-dimensional modeling, ignoring the internal structure of each component, simplifying the theoretical thermal resistance of the model, and calculating according to the 45° heat conduction model; Step 2.2, in three-dimensional modeling, ignoring the part with a thickness much smaller than other devices, converting it into a surface boundary condition, and setting the corresponding physical properties.

3. The thermal analysis method of high-heat SiC MOSFET power devices based on finite element method according to claim 1, wherein, The side length of the square is 4mm.

4. The method of claim 1, wherein the method is characterized by: The theoretical thermal resistance calculation method of the simplified model further comprises: Comparing the data sheet of the selected SiC MOSFET with the theory of the above model , it is ensured that the error is within a reasonable range.

5. The method of claim 4, wherein the method is based on a finite element method. The error is within a reasonable range, that is, the error is less than 20%.

6. The method of claim 1, wherein the method is characterized by: Step 4 comprises the following steps: Step 4.1, setting the heat convection coefficient of the contact surface exposed to air; Step 4.2, in modeling, simplifying the thin layer into a surface boundary condition, setting the corresponding material properties, and setting the heat convection coefficient of the thin layer; Step 4.3, simplifying the heat sink model, and replacing it with an equivalent increase in the heat convection coefficient of the area where the heat sink is located.

7. The high heat consumption SiC MOSFET power device thermal analysis method based on the finite element method according to claim 6, characterized in that, In step 4.1, the heat convection coefficient of the contact surface exposed to air is set to 5-20 ; In step 4.2, the heat convection coefficient of the thin layer is set to 0; 8. The method of claim 1, wherein the method is characterized by: The heat conduction equation is as follows: ， where is the density, is the specific heat capacity, u is the velocity vector, is the temperature gradient, is the divergence of the heat flux vector q, k is the thermal conductivity, and are the internal and external heat sources, respectively.

Images