Power device packaging heat dissipation optimization method based on three-dimensional thermal field regulation and control

Abstract

The invention relates to the technical field of power device heat dissipation, and discloses a power device packaging heat dissipation optimization method based on three-dimensional thermal field regulation and control. A power device packaging structure is divided into a heat source area, a heat dissipation substrate area and an interface material area, corresponding models are established after parameters of all the areas are determined, and a nonlinear thermodynamic simulation model is constructed. Thermal field data are obtained through transient thermal field simulation, an optimization objective function with a thermal field uniformity index, maximum temperature rise and thermal resistance as indexes is defined, and heat dissipation substrate parameters and interface material heat conductivity coefficient distribution are optimized through an improved particle swarm optimization algorithm. And performing sensitivity analysis by using an adjoint variable method, and dynamically adjusting model parameters according to whether the uniformity index of the thermal field reaches the standard or not. The method can accurately simulate the thermal field, optimize the heat dissipation performance in a multi-index manner, improve the optimization efficiency, effectively solve the packaging heat dissipation problem of the power device, and improve the performance and reliability of the power device.

Description

Technical Field

[0001] The present invention relates to the technical field of power device heat dissipation, and in particular to a power device packaging heat dissipation optimization method based on three-dimensional thermal field regulation. Background Art

[0002] In modern electronics, power devices are widely used in key areas such as power electronics, new energy, and automotive electronics. From converters in high-voltage transmission systems to motor controllers in electric vehicles, power devices play an indispensable role. As power devices move toward higher power density and miniaturization, heat dissipation becomes a key constraint on performance and reliability.

[0003] Traditional power device packaging heat dissipation methods, which rely on simple heat sinks or air cooling structures, are unable to meet the growing heat dissipation needs. During high-power operation, power chips generate a large amount of heat. If this heat cannot be dissipated promptly and effectively, the chip temperature will rise sharply. Excessive temperatures not only degrade power device performance, such as increased electron migration and reduced carrier mobility, but also accelerate device aging and failure, seriously affecting system reliability and service life.

[0004] Existing heat dissipation technologies have many limitations. On the one hand, there is often a lack of methods for fine-tuning the structural parameters of heat dissipation substrates. Improper settings of the height, thickness, and spacing of the fins result in insufficient utilization of the heat dissipation area or excessive air flow resistance, making it impossible to achieve efficient heat dissipation. On the other hand, when processing interface materials, their multi-physical field characteristics under temperature changes are ignored. The thermal conductivity and contact thermal resistance of interface materials will change with the temperature gradient, and traditional methods do not fully consider these factors, resulting in a large thermal resistance during heat transfer at the interface, affecting the overall heat dissipation effect.

[0005] Furthermore, existing models for thermal field simulation struggle to accurately reflect the actual heat transfer within power device packaging structures under complex operating conditions. The transient heating characteristics of heat sources and the thermal conduction coupling between heat sink substrates and interface materials are not fully and accurately simulated, rendering heat dissipation designs based on these models ineffective in solving practical problems.

[0006] With the rapid development of emerging technologies such as 5G communications and artificial intelligence, performance requirements for power devices are increasing, and heat dissipation issues are becoming increasingly prominent. For example, the power amplifiers in 5G base stations need to dissipate a large amount of heat within a confined space, and traditional heat dissipation technologies are no longer able to meet these requirements. Therefore, it is urgent to develop an efficient heat dissipation optimization method for power device packaging based on three-dimensional thermal field control to overcome the bottleneck of existing heat dissipation technology and improve the overall performance of power devices. Summary of the Invention

[0007] The object of the present invention is to provide a power device package heat dissipation optimization method based on three-dimensional thermal field regulation to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing heat dissipation of power device packaging based on three-dimensional thermal field control, the method comprising:

[0009] Step S1: Divide the power device packaging structure into a heat source area, a heat dissipation substrate area, and an interface material area, and determine the thermophysical parameters and geometric parameters of each area;

[0010] Step S2: Establishing a three-dimensional thermal field model of the heat source area, a heat conduction model of the heat dissipation substrate area, and a multi-physics field coupling model of the interface material, respectively. Connecting the models according to the actual heat conduction path, and then constructing a nonlinear thermodynamic simulation model of the packaging structure;

[0011] Step S3: Performing transient thermal field simulation on the nonlinear thermodynamic simulation model to obtain thermal field distribution and temperature gradient data of the packaging structure;

[0012] Step S4: defining a heat dissipation optimization objective function, with thermal field uniformity index, maximum temperature rise, and thermal resistance as optimization indicators;

[0013] Step S5: Using an intelligent optimization algorithm to optimize the parameters of the fin height, thickness, and spacing in the heat dissipation substrate area, and adjusting the thermal conductivity distribution of the interface material to generate an initial optimization solution set;

[0014] Step S6: Determine whether the thermal field uniformity index meets the threshold condition; if not, update the model parameters in step S2 based on the sensitivity analysis of the heat conduction path, and return to step S5; if satisfied, use the optimized parameters as the heat dissipation design benchmark for the packaging structure.

[0015] Preferably, the heat source area includes power chips, bonding wires and solder joints; the heat dissipation substrate area includes a metal substrate, a fin array and a microchannel structure; the interface material area includes a thermal conductive adhesive, a phase change material and a thermal interface composite layer;

[0016] The thermophysical parameters include the thermal conductivity, specific heat capacity and thermal expansion coefficient of the material; the geometric parameters include the height and spacing of the fins, the width and depth of the microchannels, and the thickness distribution of the interface material.

[0017] Preferably, the three-dimensional thermal field model is solved by discretization using the finite volume method, and a heat source transient power density function is introduced;

[0018] The multi-physics coupling model includes a thermal-stress coupling equation for simulating the deformation of the interface material and the change of contact thermal resistance under temperature gradient;

[0019] The heat conduction model is a steady-state and transient hybrid model based on Fourier's law, combined with the convection boundary conditions on the surface of the heat dissipation substrate.

[0020] Preferably, in step S4, the thermal field uniformity index is defined by calculating the ratio of the standard deviation of the temperature distribution to the average temperature rise; the objective function expression is:

[0021] min(α·σ T +β·T max +γ·R th )

[0022] Among them, σ T is the temperature standard deviation, T max is the maximum temperature rise, R th is the total thermal resistance, and α, β, and γ are weighting coefficients.

[0023] Preferably, in step S5, the intelligent optimization algorithm is an improved particle swarm optimization algorithm, and its fitness function is the objective function value defined in step S4; the optimization variables include the discretized combination of fin height and spacing and the gradient thermal conductivity distribution of the interface material;

[0024] The initial optimization solution set is screened through the Pareto front to retain the compromise solution between thermal resistance and temperature rise.

[0025] Preferably, in step S6, the sensitivity analysis adopts the adjoint variable method to calculate the partial derivatives of the thermal field uniformity index with respect to each geometric parameter, and updates the fin spacing and the interface material thickness based on the gradient descent strategy.

[0026] Preferably, in the multi-physics field coupling model, the thermal-stress coupling equation is realized by iteratively solving the thermoelastic control equation, wherein the Young's modulus of the interface material is nonlinearly related to temperature, and its constitutive equation is fitted by experimental data.

[0027] Preferably, the improved particle swarm optimization algorithm introduces an inertia weight adaptive adjustment mechanism, and its weight coefficient changes dynamically with the number of iterations, and the expression is:

[0028]

[0029] Among them, w(k) is the inertia weight at the kth iteration, k is the current iteration number, K is the total number of iterations, and w max With w min are the upper and lower limits of weight respectively.

[0030] Preferably, in step S3, the transient thermal field simulation adopts an explicit time integration method, combined with a variable step control strategy, and automatically shortens the time step when the temperature change rate exceeds a threshold.

[0031] Preferably, in the gradient descent strategy, the update step size is proportional to the absolute value of the partial derivative of the sensitivity analysis, and a maximum step size constraint is set to prevent parameter oscillation.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] The present invention divides the power device packaging structure into a heat source area, a heat dissipation substrate area, and an interface material area, and determines the precise thermal physical parameters and geometric parameters of each area. On this basis, a three-dimensional thermal field model, a heat conduction model, and a multi-physics field coupling model are respectively established. Each model is connected according to the actual heat conduction path to construct a nonlinear thermodynamic simulation model. This makes it possible to comprehensively and accurately simulate the heat transfer process of the packaging structure under various working conditions and obtain accurate thermal field distribution and temperature gradient data. Compared with traditional simulation methods, it can more realistically reflect the complex thermal phenomena inside the power device and provide a reliable basis for subsequent heat dissipation optimization.

[0034] A heat dissipation optimization objective function is defined, using the thermal field uniformity index, maximum temperature rise, and thermal resistance as optimization indicators. The introduction of the thermal field uniformity index allows for a quantitative assessment of the uniformity of the thermal field distribution. By optimizing this index, local overheating of power devices can be effectively avoided. Taking both maximum temperature rise and thermal resistance into account simultaneously, the optimization direction can be flexibly adjusted according to actual needs in different application scenarios. An intelligent optimization algorithm is used to optimize the fin height, thickness, and spacing in the heat dissipation substrate area, and the thermal conductivity distribution of the interface material is adjusted. This ensures thermal field uniformity while reducing the maximum temperature rise and thermal resistance, significantly improving overall heat dissipation performance.

[0035] An improved particle swarm optimization algorithm is employed, which incorporates an adaptive inertia weight adjustment mechanism, where the weight coefficient changes dynamically with the number of iterations. In the initial optimization phase, a larger inertia weight enables particles to quickly search a wider range of the solution space, improving global search capabilities. As iterations progress, the inertia weight decreases, allowing particles to focus on local areas for a refined search, improving search accuracy. This adaptive adjustment mechanism makes the optimization process more efficient, enabling faster identification of the optimal parameter combination that meets cooling requirements. Compared to traditional optimization algorithms, this significantly shortens optimization time and improves design efficiency.

[0036] To determine whether the thermal uniformity index meets the threshold conditions, a sensitivity analysis of the heat conduction path is performed based on the adjoint variable method, calculating the partial derivatives of the thermal uniformity index with respect to various geometric parameters. Based on these partial derivatives, a gradient descent strategy is used to update model parameters, such as fin spacing and interface material thickness. This dynamic adjustment mechanism based on sensitivity analysis enables targeted optimization of key parameters, making the optimization process more scientific and rational, further improving the accuracy and effectiveness of the thermal design, and ensuring that the final thermal design benchmark achieves optimal heat dissipation.

[0037] In transient thermal field simulations, an explicit time integration method combined with a variable step-size control strategy is employed. The time step is automatically shortened when the temperature change rate exceeds a threshold. This accurately captures the rapid temperature changes of power devices during operation and avoids simulation errors caused by excessive time steps. This strategy ensures that simulation results truly reflect the thermal characteristics of power devices in actual operation, providing strong support for evaluating their reliability under various operating conditions. It helps to proactively identify and optimize potential heat dissipation issues, thereby improving the stability and service life of power devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a working principle diagram of the power device packaging heat dissipation optimization method based on three-dimensional thermal field control according to the present invention;

[0039] Figure 2 Workflow diagram for determining the structural area and parameters of power device packaging;

[0040] Figure 3 Workflow diagram for parameter optimization and solution generation;

[0041] Figure 4 This is the workflow diagram for thermal field uniformity index judgment and parameter update. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0043] See also Figure 1-4 The present invention provides a power device packaging heat dissipation optimization method based on three-dimensional thermal field control, and its overall implementation scheme is as follows:

[0044] Step S1: Divide the power device packaging structure into a heat source area, a heat dissipation substrate area, and an interface material area, and determine the thermophysical parameters and geometric parameters of each area.

[0045] Step S2: Establish a three-dimensional thermal field model of the heat source area, a heat conduction model of the heat dissipation substrate area, and a multi-physics field coupling model of the interface material respectively. Connect the models according to the actual heat conduction path to construct a nonlinear thermodynamic simulation model of the packaging structure.

[0046] Step S3: performing transient thermal field simulation on the nonlinear thermodynamic simulation model to obtain thermal field distribution and temperature gradient data of the packaging structure.

[0047] Step S4: defining a heat dissipation optimization objective function, with the thermal field uniformity index, maximum temperature rise, and thermal resistance as optimization indicators.

[0048] Step S5: Using an intelligent optimization algorithm to optimize the parameters of the fin height, thickness and spacing in the heat dissipation substrate area, and adjusting the thermal conductivity distribution of the interface material to generate an initial optimization solution set.

[0049] Step S6: Determine whether the thermal field uniformity index meets the threshold condition; if not, update the model parameters in step S2 based on the sensitivity analysis of the heat conduction path, and return to step S5; if satisfied, use the optimized parameters as the heat dissipation design benchmark for the packaging structure.

[0050] The implementation of the present invention will be further described below with reference to Examples 1 to 5.

[0051] Example 1:

[0052] This embodiment further clarifies the specific composition of each region of the power device packaging structure and the detailed contents of the thermophysical parameters and geometric parameters, providing an accurate data basis for subsequent model building and heat dissipation optimization.

[0053] In this embodiment, the various regions of the power device packaging structure mentioned in step S1 are described in detail. The heat source area includes the power chip, bonding wires, and solder joints. The power chip is the core component that generates heat and continuously releases a large amount of heat energy during operation. The bonding wires are used to connect the power chip to other components, and their material and structure affect the efficiency of heat transfer. The solder joints play a role in fixing and electrically connecting, and also participate in heat conduction.

[0054] The heat dissipation substrate area includes a metal substrate, fin array, and microchannel structure. As the primary heat dissipation carrier, the metal substrate requires excellent thermal conductivity, and common metal materials such as copper and aluminum are widely used. The fin array improves heat dissipation efficiency by increasing the heat dissipation area, and its geometry and layout have a significant impact on the heat dissipation effect. The microchannel structure is a tiny channel designed within the substrate. The flow of fluid in the channel removes heat, achieving efficient heat dissipation.

[0055] The interface material area includes thermally conductive adhesive, phase change material, and thermal interface composite layer. Thermally conductive adhesive fills the tiny gap between the heat source and the heat sink, enhancing heat transfer. Phase change material undergoes a phase change when the temperature rises, absorbing a large amount of heat and thus regulating the temperature. The thermal interface composite layer integrates the properties of multiple materials to further optimize the thermal conductivity of the interface.

[0056] In terms of thermophysical parameters, a material's thermal conductivity determines how quickly heat is conducted through the material. The higher the thermal conductivity, the better the heat dissipation. For example, copper has a much higher thermal conductivity than typical plastics. The specific heat capacity reflects how easily a material absorbs heat and changes temperature. The thermal expansion coefficient relates to the dimensional stability of different materials under temperature fluctuations. These parameters are crucial for accurately simulating heat transfer processes.

[0057] Among the geometric parameters, the height and spacing of the fins directly affect the heat dissipation area and air flow resistance. Appropriately increasing the fin height can increase the heat dissipation area, but fins that are too high may result in poor air flow and reduce heat dissipation efficiency. Fin spacing that is too small increases air flow resistance, while too large reduces the heat dissipation area. The width and depth of the microchannels affect the flow rate and heat exchange efficiency of the fluid. Properly designing the microchannel dimensions can achieve optimal heat dissipation. The thickness distribution of the interface material also affects heat transfer. Excessively thick interface material may increase thermal resistance, while too thin an interface material may fail to effectively fill the gap.

[0058] Example:

[0059] This embodiment describes in detail the establishment method and solution method of each model in step S2, as well as the implementation process of the thermal-stress coupling equation in the multi-physics field coupling model, to ensure that the established nonlinear thermodynamic simulation model can accurately simulate actual heat transfer and physical changes.

[0060] In step S2, the finite volume method is used to discretize and solve the three-dimensional thermal field model of the heat source area. The finite volume method divides the solution area into a series of control volumes and integrates the conservation equations within each control volume to obtain a discretized set of equations. To more accurately simulate the heat generation of the heat source, a heat source transient power density function is introduced. This function is defined based on the power variation of the power device under different operating conditions and can reflect the change of the heat source's heat intensity over time in real time.

[0061] The heat conduction model for the heat sink region is a hybrid steady-state and transient model based on Fourier's law. Fourier's law describes the conduction of heat through a medium. The steady-state model is suitable for situations where the temperature distribution of the heat sink is relatively stable, while the transient model captures temperature variations over time. Combined with the convection boundary conditions on the heat sink surface, the heat exchange between the heat sink and the surrounding environment is considered, and the intensity of this heat exchange is described by the convective heat transfer coefficient.

[0062] The multi-physics coupling model of the interface material includes a thermal-stress coupling equation, which is used to simulate the deformation and contact thermal resistance changes of the interface material under temperature gradients. The thermal-stress coupling equation is implemented by iteratively solving the thermoelastic control equation. In this process, the Young's modulus of the interface material has a nonlinear relationship with temperature. In order to accurately describe this relationship, its constitutive equation is fitted by experimental data. For example, the mechanical properties of the interface material at different temperatures are tested, and the data points of the Young's modulus changing with temperature are obtained. Then, a suitable mathematical method is used to fit the data to obtain an expression that can accurately reflect the relationship between Young's modulus and temperature. In this way, during the simulation process, the influence of temperature on the mechanical properties of the interface material can be more realistically considered, thereby more accurately simulating the changes in contact thermal resistance.

[0063] Example 3:

[0064] This embodiment details the definition and calculation of the heat dissipation optimization objective function and introduces the specific implementation details of the intelligent optimization algorithm (improved particle swarm optimization algorithm), including the setting of the fitness function and the selection of optimization variables, as well as how to screen the initial optimization solution set through the Pareto front.

[0065] In step S4, the thermal uniformity index (THI) is defined by calculating the ratio of the standard deviation of the temperature distribution to the average temperature rise. This index intuitively reflects the uniformity of the thermal field distribution within the power device package structure. The standard deviation of the temperature distribution reflects the degree of dispersion of the temperature at each point relative to the average temperature, while the average temperature rise indicates the overall temperature increase. The smaller the ratio of the two, the more uniform the thermal field distribution.

[0066] The objective function expression is:

[0067] min(α·σ T +β·T max +γ·R th )

[0068] Among them, σ T is the temperature standard deviation, T max is the maximum temperature rise, R th is the total thermal resistance, and α, β, and γ are weighting coefficients. These weighting coefficients are set based on actual needs to adjust the importance of different optimization indicators in the objective function. For example, if thermal field uniformity is more important, the value of α can be appropriately increased; if the maximum temperature rise limit is more stringent, the value of β can be increased.

[0069] In step S5, the intelligent optimization algorithm used is an improved particle swarm optimization algorithm. The fitness function of this algorithm is the objective function value defined in step S4. The optimization variables include the discretized combination of fin height and spacing and the gradient thermal conductivity distribution of the interface material. By discretizing the fin height and spacing, the optimal combination can be found within a limited range of values. For the gradient thermal conductivity distribution of the interface material, considering the different requirements for thermal conductivity performance at different locations, the overall heat dissipation effect is improved by optimizing the distribution of thermal conductivity.

[0070] The improved particle swarm optimization algorithm introduces an inertia weight adaptive adjustment mechanism, and its weight coefficient changes dynamically with the number of iterations. The expression is: Among them, w(k) is the inertia weight at the kth iteration, k is the current iteration number, K is the total number of iterations, and w max With w min In the early stages of the algorithm, a larger inertia weight helps particles search in a larger range, speeding up the global search. As the number of iterations increases, the inertia weight gradually decreases, causing particles to focus more on local search and improving search accuracy.

[0071] The initial set of optimized solutions is filtered through the Pareto front to retain compromise solutions between thermal resistance and temperature rise. The Pareto front refers to a set of solutions in a multi-objective optimization problem where no objective can be further improved by sacrificing other objectives. In this invention, thermal resistance and temperature rise are two interrelated optimization objectives. Through Pareto front screening, solutions that achieve a good balance between thermal resistance and temperature rise can be obtained, avoiding excessive pursuit of one metric while neglecting others.

[0072] Example 4:

[0073] This embodiment describes in detail the sensitivity analysis method (adjoint variable method) in step S6 and the specific process of updating the model parameters based on the gradient descent strategy, to ensure that when the thermal field uniformity index does not meet the threshold condition, the model parameters can be effectively optimized and the heat dissipation effect can be gradually improved.

[0074] In step S6, the sensitivity analysis adopts the adjoint variable method. The adjoint variable method is an efficient method for calculating the partial derivatives of a multivariable function with respect to multiple independent variables. In the present invention, the partial derivatives of the thermal field uniformity index with respect to each geometric parameter are calculated by the adjoint variable method. These partial derivatives reflect the degree of influence of a small change in each geometric parameter on the thermal field uniformity index. For example, the partial derivative of the thermal field uniformity index with respect to the fin spacing is calculated. If the absolute value of the partial derivative is large, it means that the change in the fin spacing has a more significant effect on the thermal field uniformity; conversely, if the absolute value of the partial derivative is small, the effect is small.

[0075] The fin spacing and interface material thickness are updated based on the gradient descent strategy. In the gradient descent strategy, the update step size is proportional to the absolute value of the partial derivative of the sensitivity analysis. This means that geometric parameters that have a greater impact on the thermal field uniformity index will be updated with a larger step size so that they can be adjusted in the optimization direction more quickly; while geometric parameters that have a smaller impact on the thermal field uniformity index will be updated with a smaller step size to avoid parameter oscillation caused by excessive adjustment. At the same time, a maximum step size constraint is set to prevent parameter oscillation. If the update step size is too large, the parameters may jump back and forth near the optimal solution and fail to converge to the optimal value. By setting the maximum step size, the amplitude of each parameter update is limited, ensuring the stability of the parameter adjustment.

[0076] In practice, the fin spacing and interface material thickness are updated based on the calculated partial derivatives and the set maximum step size. The updated parameters are then substituted into the model in step S2, the nonlinear thermodynamic simulation model is rebuilt, and the optimization process returns to step S5 to continue until the thermal field uniformity index meets the threshold condition.

[0077] Embodiment 5:

[0078] This embodiment details the specific implementation of the explicit time integration method and the variable step control strategy used in the transient thermal field simulation in step S3, ensuring that the change of the packaging structure temperature over time can be accurately captured during the simulation process, thereby improving the accuracy and reliability of the simulation results.

[0079] In step S3, the transient thermal field simulation uses the explicit time integration method. This method is a numerical calculation method based on a step-by-step approach. Within each time step, the temperature value at the next moment is calculated based on the temperature distribution at the current moment and the heat conduction equation. This method has the advantages of simple calculation and high efficiency, making it suitable for handling large-scale heat conduction problems.

[0080] Combined with a variable step-size control strategy, the time step is automatically shortened when the temperature change rate exceeds a threshold. The temperature change rate reflects how quickly the temperature changes over time. During the operation of a power device package structure, rapid temperature changes may occur, such as at the moment the power device is turned on or off. If a fixed time step is used for simulation at this time, the simulation results may be inaccurate and unable to capture the rapid temperature changes. Through the variable step-size control strategy, when the temperature change rate is detected to exceed a pre-set threshold, the time step is automatically reduced. This improves the time resolution of the simulation in areas with drastic temperature changes and more accurately calculates the temperature change process. Conversely, when the temperature changes are relatively gentle, the time step can be appropriately increased to improve computational efficiency and reduce calculation time.

[0081] During the actual simulation process, the initial time step and temperature change rate threshold are first set. After each time step is calculated, the temperature change rate within the current time step is calculated. If the temperature change rate exceeds the threshold, the time step is shortened according to a specific rule and the temperature distribution within that time step is recalculated. If the temperature change rate does not exceed the threshold, the time step can be appropriately increased. By continuously adjusting the time step, accurate simulation of the transient thermal field of the package structure is achieved, resulting in more reliable thermal field distribution and temperature gradient data.

[0082] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0083] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing heat dissipation of power device packaging based on three-dimensional thermal field control, characterized in that: The following steps are involved: Step S1: Divide the power device packaging structure into a heat source area, a heat dissipation substrate area, and an interface material area, and determine the thermophysical parameters and geometric parameters of each area; Step S2: Establishing a three-dimensional thermal field model of the heat source area, a heat conduction model of the heat dissipation substrate area, and a multi-physics field coupling model of the interface material, respectively. Connecting the models according to the actual heat conduction path, and then constructing a nonlinear thermodynamic simulation model of the packaging structure; Step S3: Performing transient thermal field simulation on the nonlinear thermodynamic simulation model to obtain thermal field distribution and temperature gradient data of the packaging structure; Step S4: defining a heat dissipation optimization objective function, with thermal field uniformity index, maximum temperature rise, and thermal resistance as optimization indicators; Step S5: Using an intelligent optimization algorithm to optimize the parameters of the fin height, thickness, and spacing in the heat dissipation substrate area, and adjusting the thermal conductivity distribution of the interface material to generate an initial optimization solution set; Step S6: determining whether the thermal field uniformity index meets a threshold condition; If not, the model parameters in step S2 are updated based on the sensitivity analysis of the heat conduction path, and the process returns to step S5; If satisfied, the optimized parameters are used as the heat dissipation design benchmark for the packaging structure.

2. The method according to claim 1, wherein: In step S1, the heat source area includes power chips, bonding wires and solder joints; the heat dissipation substrate area includes a metal substrate, a fin array and a microchannel structure; and the interface material area includes a thermal conductive adhesive, a phase change material and a thermal interface composite layer. The thermophysical parameters include the thermal conductivity, specific heat capacity and thermal expansion coefficient of the material; the geometric parameters include the height and spacing of the fins, the width and depth of the microchannels, and the thickness distribution of the interface material.

3. The method according to claim 1, wherein: In step S2, the three-dimensional thermal field model is discretized and solved using the finite volume method, and a heat source transient power density function is introduced; The multi-physics coupling model includes a thermal-stress coupling equation for simulating the deformation of the interface material and the change of contact thermal resistance under temperature gradient; The heat conduction model is a steady-state and transient hybrid model based on Fourier's law, combined with the convection boundary conditions on the surface of the heat dissipation substrate.

4. The method according to claim 1, wherein: In step S4, the thermal field uniformity index is defined by calculating the ratio of the standard deviation of the temperature distribution to the average temperature rise; the objective function expression is: min(a·s T +β·T max +γ·R th ) Among them, σ T is the temperature standard deviation, T max is the maximum temperature rise, R th is the total thermal resistance, and α, β, and γ are weighting coefficients.

5. The method according to claim 1, wherein: In step S5, the intelligent optimization algorithm is an improved particle swarm optimization algorithm, and its fitness function is the objective function value defined in step S4; the optimization variables include the discretized combination of fin height and spacing and the gradient thermal conductivity distribution of the interface material; The initial optimization solution set is screened through the Pareto front to retain the compromise solution between thermal resistance and temperature rise.

6. The method according to claim 1, wherein: In step S6, the sensitivity analysis adopts the adjoint variable method to calculate the partial derivatives of the thermal field uniformity index with respect to each geometric parameter, and updates the fin spacing and the interface material thickness based on the gradient descent strategy.

7. The method according to claim 3, wherein: In the multi-physics field coupling model, the thermal-stress coupling equation is implemented by iteratively solving the thermoelastic control equation, wherein the Young's modulus of the interface material is nonlinearly related to temperature, and its constitutive equation is fitted by experimental data.

8. The method according to claim 5, characterized in that: The improved particle swarm optimization algorithm introduces an inertia weight adaptive adjustment mechanism, and its weight coefficient changes dynamically with the number of iterations. The expression is: Among them, w(k) is the inertia weight at the kth iteration, k is the current iteration number, K is the total number of iterations, and w max With w min are the upper and lower limits of weight respectively.

9. The method according to claim 1, wherein: In step S3, the transient thermal field simulation adopts an explicit time integration method combined with a variable step control strategy to automatically shorten the time step when the temperature change rate exceeds a threshold.

10. The method according to claim 6, wherein: In the gradient descent strategy, the update step size is proportional to the absolute value of the partial derivative of the sensitivity analysis, and a maximum step size constraint is set to prevent parameter oscillation.

Images