Battery pack heat dissipation structure intelligent design method and device, and storage medium

Abstract

The application discloses a battery pack heat dissipation structure intelligent design method and device and a storage medium. The method comprises the following steps: determining a heat dissipation structure design domain, dividing a grid, and setting a design variable; loading a load on a current structure according to an assembly scheme, solving a contact pressure distribution, and obtaining an interface thermal resistance distribution according to a pressure-thermal resistance characteristic of a heat conduction interface material; performing thermal analysis by taking the interface thermal resistance distribution as a boundary condition, and obtaining a power device temperature peak value; calculating a sensitivity of the temperature peak value to each design variable, wherein a direct influence caused by a material thermal conductivity change and an indirect influence caused by a structure stiffness change through a contact pressure and an interface thermal resistance change are simultaneously considered; iteratively updating the design variable under a constraint condition according to the sensitivity until a convergence condition is met, and outputting a heat dissipation structure geometric shape. The application can comprehensively consider an influence of an assembly contact state on heat transfer performance in heat dissipation structure optimization, and improves accuracy of heat dissipation structure design.

Description

Technical Field

[0001] This application relates to the field of battery pack heat dissipation structure simulation design technology, specifically to a battery pack heat dissipation structure intelligent design method, device and storage medium. Background Technology

[0002] As household cleaning devices such as robotic vacuum cleaners and mops evolve towards miniaturization, high power, and fast charging, the charge / discharge rates and power densities that battery packs must withstand within limited spaces are constantly increasing. Power devices such as MOSFETs on the protection board generate concentrated heat when operating at high currents. This heat is typically transferred to the battery pack casing using aluminum alloy heat dissipation structures and thermally conductive interface materials. Existing heat dissipation structure designs employ methods such as finite element thermal analysis, parameter optimization, or topology optimization to reduce the maximum temperature of the heat source or the overall thermal resistance, while balancing heat dissipation performance and structural weight.

[0003] The actual heat transfer capacity between the heat dissipation structure and power devices is affected not only by the thermal conductivity of the materials but also by the assembly method (such as screws and clips) and local contact pressure. Changes in the pressure applied to the thermal interface material cause changes in interface thermal resistance, and fastening positions, structural stiffness, and assembly deviations can also lead to uneven contact pressure distribution. Current designs that simplify interface thermal resistance to a fixed value or perform thermal analysis based solely on ideal contact conditions fail to reflect the interplay between structural geometric changes and contact pressure and interface thermal resistance, easily resulting in discrepancies between simulation results and actual temperature rise after assembly. Therefore, how to reasonably consider the relationship between assembly stress, contact heat transfer, and structural geometry in heat dissipation structure design remains a problem that needs to be solved to improve design accuracy and engineering applicability. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this specification provides a method, device and storage medium for intelligent design of battery pack heat dissipation structure.

[0005] According to a first aspect of the embodiments of this application, a smart design method for a battery pack heat dissipation structure is provided, including: The design domain of the heat dissipation structure is determined based on the internal component layout of the battery pack. The design domain is then divided into a grid, and design variables are assigned to each unit. Based on the heat dissipation structure assembly scheme, apply assembly load to the structure corresponding to the current design variables and solve the structural equilibrium equation to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. Based on the pressure-thermal resistance characteristics of the thermally conductive interface material, the contact pressure distribution is mapped to the interface thermal resistance distribution at each position of the contact surface. Thermal analysis was performed on the heat dissipation structure using the interface thermal resistance distribution as a boundary condition to obtain the peak temperature of the power device. Calculate the sensitivity of the peak temperature to each design variable. The sensitivity includes a direct component and an indirect component. The direct component corresponds to the influence of the design variable on the temperature field through the change in the thermal conductivity of the material. The indirect component corresponds to the influence of the design variable on the temperature field through the change in the contact pressure distribution and the interface thermal resistance distribution caused by the change in the structural stiffness. Update each design variable under constraints based on the sensitivity; repeat the above steps of solving the structural equilibrium equations and updating each design variable until the convergence condition is met, and output the converged heat dissipation structure geometry.

[0006] As an optional approach, the step of applying assembly loads to the structure corresponding to the current design variables and solving the structural equilibrium equations to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device includes: calculating the axial preload at each fastening point based on the preload torque and fastener specification parameters of each fastener of the heat dissipation structure; setting up normal contact pairs between the heat dissipation structure and the power device; assembling the stiffness matrix using the axial preload as a load and the equivalent elastic modulus corresponding to each unit design variable and solving the displacement field; and extracting the normal contact pressure at each node of the contact surface from the displacement field to form the contact pressure distribution.

[0007] As an optional approach, mapping the contact pressure distribution to the interfacial thermal resistance distribution at various locations on the contact surface includes: fitting measured thermal resistance data of the thermally conductive interface material under different pressures to obtain a monotonically decreasing mapping function with pressure as the independent variable and area specific thermal resistance as the dependent variable; substituting the contact pressure values ​​at various locations on the contact surface into the monotonically decreasing mapping function to calculate the area specific thermal resistance value at each location; and setting the area specific thermal resistance of locations with contact pressures lower than a preset minimum effective pressure threshold as a cutoff upper limit value in a smooth transition manner.

[0008] As an optional approach, the monotonically decreasing mapping function adopts a power-law form, using the nominal thermal resistance value under the reference pressure and the reference pressure as reference parameters, and the negative power of the ratio of the local contact pressure to the reference pressure as a scaling factor; the power exponent is determined by least-squares fitting of the measured thermal resistance data of the thermally conductive interface material at multiple pressure points.

[0009] As an optional approach, the thermal analysis is a transient thermal analysis covering the entire charging cycle; the heat source load is calculated based on the charging current curve and the on-resistance of the power device at the operating temperature; the temperature peak is the highest temperature reached by the power device during the entire charging cycle; and the time of occurrence of the temperature peak and the corresponding temperature field distribution are recorded simultaneously with obtaining the temperature peak.

[0010] As an optional approach, the calculation of the indirect components includes: solving the thermal adjoint equation based on the temperature field distribution at the moment the temperature peak occurs to obtain the thermal adjoint vector; calculating the sensitivity of the temperature peak to the interfacial thermal resistance of each node on the contact surface based on the thermal adjoint vector; multiplying the sensitivity of the temperature peak to the interfacial thermal resistance of each node by the sensitivity of the interfacial thermal resistance at each node to the contact pressure node by node to obtain the comprehensive weight of each node; combining the comprehensive weights of each node to form an equivalent force vector, which is used as the right-hand side term of the structural adjoint equation to obtain the structural adjoint vector; and calculating the indirect components of each design variable based on the structural adjoint vector and the partial derivatives of the stiffness matrices of each element with respect to the design variables.

[0011] As an optional approach, the sensitivity of the interfacial thermal resistance at each node to the contact pressure is obtained by differentiating the monotonically decreasing mapping function with respect to pressure, and its value is negative; the equivalent force vector combines the comprehensive weights of all nodes on the contact surface into a single right-hand term, compressing the multiple structural adjoint equations that originally needed to be solved separately for each contact node into a single solution.

[0012] As an optional approach, the equivalent elastic modulus and equivalent thermal conductivity of each unit are determined by the penalty interpolation method for solid isotropic materials, which is obtained by multiplying the penalty power of the design variables of each unit with the corresponding property value of the matrix material; the penalty power is a preset constant with a value range of 2 to 4.

[0013] According to a second aspect of the embodiments of this application, a smart design device for a battery pack heat dissipation structure is also provided, comprising: The design domain construction module is used to determine the design domain of the heat dissipation structure based on the layout of the internal components of the battery pack, divide the design domain into a grid and assign design variables to each unit. The contact pressure calculation module is used to apply assembly loads to the structure corresponding to the current design variables and solve the structural equilibrium equations according to the heat dissipation structure assembly scheme, so as to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. The interface thermal resistance mapping module is used to map the contact pressure distribution to the interface thermal resistance distribution at each position of the contact surface based on the pressure-thermal resistance characteristics of the thermally conductive interface material. The thermal analysis module is used to perform thermal analysis on the heat dissipation structure with the interface thermal resistance distribution as the boundary condition to obtain the temperature peak of the power device. The sensitivity calculation module is used to calculate the sensitivity of the temperature peak to each design variable. The sensitivity includes a direct component and an indirect component. The direct component corresponds to the influence of the design variable on the temperature field caused by the change in the thermal conductivity of the material. The indirect component corresponds to the influence of the design variable on the temperature field caused by the change in the contact pressure distribution and the interface thermal resistance distribution due to the change in the structural stiffness. The design variable update and iteration control module is used to update each design variable under constraints according to the sensitivity; repeat the above steps of solving the structural equilibrium equation to update each design variable until the convergence condition is met, and output the converged heat dissipation structure geometry.

[0014] According to a third aspect of the embodiments of this application, a computer-readable storage medium is provided, which, when an executable computer program in the storage medium is executed by a processor, enables the implementation of the method described in the first aspect.

[0015] This application incorporates the influence of actual assembly conditions on interfacial heat transfer performance during the optimization of the heat dissipation structure. Based on the contact pressure distribution formed under the current structure's fastening load, and combined with the pressure-thermal resistance characteristics of the thermally conductive interface material, the spatially varying interfacial thermal resistance is determined and used as a boundary condition for thermal analysis. This establishes a continuous correlation between structural geometry, structural stiffness, contact pressure, interfacial thermal resistance, and power device temperature. The direct effects of design variables on the temperature field through material thermal conductivity, as well as the indirect effects through structural stiffness, contact pressure, and interfacial thermal resistance, are uniformly incorporated into sensitivity calculations, and the structure is iteratively updated accordingly. This avoids design deviations caused by fixed interfacial thermal resistance assumptions, enabling the optimized structure to proactively improve the contact thermal conductivity of the power device region with limited material usage, reduce temperature peaks, and improve the consistency between predicted heat dissipation performance and actual assembly conditions. This enhances the engineering applicability and stability of the heat dissipation structure design results under mass production conditions. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this specification and, together with the description, serve to explain the principles of this specification.

[0017] Figure 1 This is a flowchart of an intelligent design method for a battery pack heat dissipation structure according to one embodiment of this application.

[0018] Figure 2 This is a schematic diagram of the contact pressure distribution calculation process in one embodiment of this application.

[0019] Figure 3 This is a schematic diagram illustrating the coupling relationship between contact pressure, interfacial thermal resistance, and temperature field in one embodiment of this application.

[0020] Figure 4 This is a schematic diagram of temperature peak complete sensitivity calculation in one embodiment of this application.

[0021] Figure 5 This is a schematic diagram comparing the contact thermal performance before and after optimization of the heat dissipation structure in one embodiment of this application.

[0022] Figure 6 This is a schematic block diagram of an intelligent design device for a battery pack heat dissipation structure according to this application.

[0023] Figure 7 This is a structural diagram of an electronic device according to one embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] This embodiment provides an intelligent design method for the heat dissipation structure of a battery pack, targeting the development of heat dissipation structures for high-rate charge / discharge battery packs in household cleaning equipment. Applicable devices include battery packs for robotic vacuum cleaners, robotic mops, and robotic window cleaners, with charging rates ranging from 3C to 5C. In these battery packs, parallel-connected MOSFETs on the protection board serve as the main power devices, generating highly concentrated heat during high-current charging. The heat dissipation structure is made of aluminum alloy, with one side bonded to the top surface of the MOSFET package via a thermally conductive interface material (TIM), and the other side in contact with the battery pack casing, forming a heat conduction path from the heat source to the outside. The heat dissipation structure is fastened to the casing with screws or clips, the fastening force being transmitted within the structure and generating normal pressure at the contact surface. This method runs on an engineering simulation workstation equipped with finite element analysis software, iteratively performing structural mechanics and heat transfer analyses within a topology optimization framework, automatically searching for the optimal heat dissipation structure geometry.

[0026] The implementation process of the method described in this application will be explained in detail below with reference to specific embodiments.

[0027] Please see Figure 1 , Figure 1 This is a flowchart of an intelligent design method for a battery pack heat dissipation structure according to one embodiment of this application, such as... Figure 1 As shown, the method includes steps 101-106: Step 101: Determine the design domain of the heat dissipation structure based on the internal component layout of the battery pack, divide the design domain into a grid, and assign design variables to each unit.

[0028] According to embodiments of this disclosure, using a 3D CAD assembly model of the battery pack as input, the volume of the inner cavity of the casing is extracted, and the space occupied by existing components such as the battery cells, protection PCB, connectors, and wiring harnesses is subtracted. The remaining connected volume is the design domain of the heat dissipation structure. The design domain is meshed in a finite element preprocessing environment, with hexahedral or tetrahedral elements selected as the element type. The size is determined based on the minimum feature size of the manufacturing process, for example, configured as 0.4 mm. Each mesh element is assigned a design variable. The value ranges from 0 to 1, where 1 represents that the position is completely filled with aluminum alloy material and 0 represents a cavity.

[0029] The equivalent elastic modulus and equivalent thermal conductivity of each element were determined using the solid isotropic material penalized interpolation method (SIMP), obtained by multiplying the penalty power of the design variables by the corresponding property value of the matrix material. ; ; in For unit The equivalent elastic modulus, in Pa; This is the elastic modulus of the aluminum alloy matrix, for example, 70 GPa; Equivalent thermal conductivity, in W / (m·K); The thermal conductivity of the matrix is, for example, taken as 180 W / (m·K). and These are the structural and thermal penalty powers, respectively, both dimensionless preset constants ranging from 2 to 4. The purpose of the penalty powers is to make the intermediate density ( Element performance (between 0.3 and 0.7) is significantly worse than solid performance, forcing the optimization results to converge to a clear distribution of 0 or 1. In engineering, 3 is a commonly used value, derived from the standard recommendation of the SIMP method. A penalty strength below 2 results in too many grayscale elements, while a strength above 4 makes it easy to get trapped in local optima in the early stages of optimization.

[0030] Step 102: Based on the heat dissipation structure assembly scheme, apply assembly loads to the structure corresponding to the current design variables and solve the structural equilibrium equations to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device.

[0031] According to embodiments of this disclosure, under high-rate charging conditions (3C to 5C), the heat generation power of the MOSFET increases with the square of the current, and the influence of the deviation in local thermal resistance at the contact surface on the temperature rise becomes significant. Traditional heat dissipation structure designs skip this step, assuming the interface thermal resistance is constant and directly proceeding to thermal analysis; this method first performs structural mechanics analysis in each optimization iteration to obtain the contact pressure distribution generated by the current geometry under assembly force, and this distribution changes with design variables.

[0032] Specifically, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the contact pressure distribution calculation process in one embodiment of this application. In step 1021, the axial preload of each fastening point is calculated based on the preload torque and fastener specification parameters of each fastener in the heat dissipation structure.

[0033] Taking screw fixing as an example: ; For the first Axial preload of a screw, in N; The preload torque is expressed in N·m and is specified by the assembly process specifications. The nominal diameter of the screw is in meters (m). This is the torque coefficient, which comprehensively reflects thread friction and bearing surface friction. Its value can be obtained by consulting fastener engineering handbooks or determined through torque-axial force calibration tests. The typical range for steel screws mating with aluminum alloy threaded holes is 0.15 to 0.22. For example, for an M2 screw with a torque of 0.2 N·m, take... =0.18, the axial preload is 0.2 / (0.18×0.002)≈556N.

[0034] In step 1022, a normal contact pair is set between the bottom surface of the heat dissipation structure and the top surface of the power device, and the contact type is normal hard contact. The equivalent elastic modulus corresponding to the current design variables of each unit is used. Assembly of overall stiffness matrix The external force vector is formed by the axial preload of each fastening point. Solve the structural equilibrium equations: ; This is the global nodal displacement vector, in meters (m). The dimension is N / m; The dimension is N. The solution method can be either the sparse direct decomposition method or the preconditional conjugate gradient method.

[0035] In step 1023, after the displacement field is solved, the normal contact pressure at each node of the contact surface is extracted from the normal contact pair. Taking the penalty function method as an example, the nodes... Normal contact pressure is , Penalty stiffness (unit: Pa / m) This represents the normal penetration amount. The contact pressure distribution is formed by the convergence of the normal contact pressures at all contact nodes. In terms of physical properties, the pressure is high near the screw and low at the far end; the thicker the local heat dissipation structure and the greater the bending stiffness, the more uniform the pressure distribution.

[0036] Step 103: Based on the pressure-thermal resistance characteristics of the thermally conductive interface material, map the contact pressure distribution to the interface thermal resistance distribution at each location on the contact surface.

[0037] This step converts the contact pressure distribution output from the previous steps into an area specific thermal resistance field, i.e., an interfacial thermal resistance distribution. Area specific thermal resistance refers to the thermal resistance value per unit contact area, with units of m²·K / W, reflecting the interfacial thermal conductivity. The more fully the TIM is compressed, the more internal micropores are compressed and expelled, increasing the effective heat transfer cross section, and consequently decreasing the area specific thermal resistance.

[0038] In one embodiment, measured thermal resistance data of the thermally conductive interface material under different pressures are fitted to obtain a monotonically decreasing mapping function. For example, TIM measures the area ratio thermal resistance data at several pressure points according to standard methods such as ASTM D5470. A linear regression is then performed on the logarithm of the measured data to establish a power-law form mapping function. ; For position Thermal resistance per square area, in m²·K / W; For reference pressure The supplier's nominal area thermal resistance value, in m²·K / W, for example, a silicone thermal pad has a nominal value of 2.5 × 10⁻⁶ at 50 kPa. -4 m 2 K / W; This is a reference pressure, in Pa, for example, 50 kPa; The contact pressure at this location output from the aforementioned steps, in Pa; It represents the power exponent. The determination was made by least-squares fitting of measured thermal resistance data at multiple pressure points provided by the supplier: [Data points] After taking the logarithm, we get The linear relationship is such that the absolute value of the regression slope is... Silicone-based thermal pads typically range from 0.4 to 0.8.

[0039] As an example, when the contact pressure at a certain location is 30 kPa, the area thermal resistance is... The area near the screw at 80 kPa is The difference reached 1.8 times.

[0040] For contact pressure below the preset minimum effective pressure threshold The location is such that the area-specific thermal resistance is set as the upper limit of the cutoff value in a smooth transition manner. . Obtain it from the TIM supplier's technical specifications, for example, take 10 kPa. Below this value, the TIM is not effectively compressed, and the heat conduction path is interrupted. Take the area ratio as 10 times the thermal resistance under normal operating conditions.

[0041] The smooth transition is achieved using the sigmoid function: ; in sigmoid function ; The width of the transition zone is expressed in Pa, for example, 2 kPa. Select 10% to 20%. When Much higher When the sigmoid output approaches 0, Degenerate into ;when far below When the sigmoid output approaches 1, Approaching .

[0042] Figure 3 This is a schematic diagram illustrating the coupling relationship between contact pressure, interfacial thermal resistance, and temperature field in one embodiment of this application, as shown below. Figure 3 As shown, in this embodiment, during the topology optimization process of the heat dissipation structure, a continuous coupling relationship is established between the assembly force, contact pressure, interface thermal resistance, and the temperature field of the power device. Based on the fastener preload and the current stiffness of the heat dissipation structure, the structural equilibrium equations are solved to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. Then, based on the pressure-thermal resistance characteristics of the thermally conductive interface material, the contact pressure at each location is mapped to the interface thermal resistance distribution. The regions with higher contact pressure correspond to lower interfacial thermal resistance, while the regions with lower pressure correspond to higher interfacial thermal resistance. Subsequently, using this non-uniform interfacial thermal resistance distribution as the boundary condition for thermal analysis, transient thermal analysis was performed on the power device and heat dissipation structure during the charging cycle to obtain the temperature field of the power device and extract the peak temperature. This allows differences in contact thermal conductivity caused by assembly conditions to be directly fed back into the evaluation of thermal performance.

[0043] Step 104: Perform thermal analysis on the heat dissipation structure using the interface thermal resistance distribution as the boundary condition to obtain the peak temperature of the power device.

[0044] In one embodiment, the thermal analysis is a transient thermal analysis covering the entire charging cycle, spanning from the start to the end of charging. The heat source load is determined based on the charging current curve. On-resistance of power devices at operating temperature Calculate: Instantaneous heating power . The temperature coefficient increases with increasing temperature, approximately +0.4% / °C. A nominal value of 25°C is used initially in the iterations, and subsequent iterations correct for the MOSFET temperature from previous results. The charging current curve includes a constant current phase (current remains constant at the maximum charging current) and a constant voltage phase (current gradually decreases).

[0045] The interfacial heat transfer between the bottom surface of the heat dissipation structure and the top surface of the power device is represented by the interfacial thermal resistance distribution output in the aforementioned steps. As a boundary condition, the equivalent interfacial thermal conductivity at each node is taken as . Divide the reciprocal of the product by the nominal thickness of the TIM. Apply a corresponding natural convection heat transfer boundary to the contact surface between the outer side of the heat dissipation structure and the shell. The thermal analysis model uses the same mesh as the previous steps, and the thermal conductivity of each element is the SIMP interpolated value. .

[0046] In one embodiment, the transient thermal equilibrium equations are assembled and solved: ; The node temperature vector, in K; The rate of temperature change is expressed in K / s. This is a heat capacity matrix, in J / K. The heat conduction matrix (containing each element) Contribution and interfacial thermal resistance (Contribution), unit: W / K; This is a time-varying heat load vector, in W. The time discretization uses an implicit difference scheme, with a step size of 30s to 60s during the constant current stage, and a step size of 1s to 5s near the current switching point.

[0047] After the transient thermal analysis is completed, the highest value of the power device node temperature is extracted as the temperature peak value across all time steps. Record the time when the temperature peak occurs. and the corresponding temperature field distribution . To optimize the objective function, and For use in subsequent sensitivity calculations.

[0048] Step 105: Calculate the sensitivity of the peak temperature to each design variable. In one embodiment, the sensitivity includes a direct component and an indirect component, which are added together to form complete gradient information. The direct component corresponds to the influence of the design variable on the temperature field caused by changes in the material's thermal conductivity; the indirect component corresponds to the influence of the design variable on the temperature field caused by changes in the contact pressure distribution and interface thermal resistance distribution due to changes in structural stiffness.

[0049] The calculation of the direct component employs the thermal synergistic method, specifically based on the time of temperature peak occurrence. Temperature field distribution Construct and solve the thermal adjoint equation: ; Let be the thermal adjoint vector, and the right side of the equation is the partial derivative of the objective function with respect to the nodal temperatures. For a specific node The temperature value at any given time; this vector takes a value of -1 only at the corresponding position, and 0 otherwise.

[0050] coefficient matrix of the adjoint equation Similar to the main problem (the heat conduction matrix is ​​symmetric), the matrix decomposition already completed in the main problem can be reused for direct back substitution to solve the problem, with minimal additional computational cost.

[0051] Seek The direct component is ,in Given by the SIMP interpolation rules ( ), only in the unit It has a non-zero contribution in a local range.

[0052] In one embodiment, the calculation of the indirect component is completed in three stages. The first stage involves calculating based on the thermal adjoint vector. Calculate the sensitivity of the temperature peak to the interfacial thermal resistance of each node on the contact surface, i.e. . For the heat conduction matrix pair of the first The partial derivative of the thermal resistance of an interface node is only related to the interface elements near that node.

[0053] The second section describes the sensitivity of interfacial thermal resistance at each node to contact pressure, obtained by differentiating the monotonically decreasing mapping function established in the preceding steps with respect to pressure. ; Its value is always negative, reflecting the physical relationship that thermal resistance decreases as pressure increases.

[0054] Third segment: Multiply the first and second segments node by node to obtain the nodes of the contact surface. Overall weight , characterizing the The degree of influence of minute changes in contact pressure at each node on the final temperature peak.

[0055] Based on the comprehensive weight of each node The partial derivatives of the contact forces at each node with respect to the displacement are superimposed to form an equivalent force vector, which serves as the right-hand side term of the structural adjoint equation: ; The structural adjoint vector; Given the total number of contact surface nodes, this merging process reduces the original solution required for each contact node to be calculated separately. The adjoint equation of the substructure is compressed into a single solution, that is, the equivalent force vector combines the combined weights of all nodes into a single right-hand side term. The number of nodes on the contact surface is typically in the hundreds to thousands, and this merging significantly improves computational efficiency. The solution is obtained... Then, the indirect components of each design variable are: ,in , Let be the stiffness matrix of the element in the fully filled state.

[0056] Adding the direct and indirect components yields the complete sensitivity of each design variable. Spatially, the indirect components of elements closer to the contact surface contribute significantly, as density changes in these elements simultaneously affect local heat conduction and local contact pressure; the indirect components of elements farther from the contact surface approach zero, and the sensitivity is dominated by the direct components.

[0057] Figure 4 This is a schematic diagram illustrating the calculation of the complete sensitivity of the temperature peak in one embodiment of this application. For example... Figure 4 As shown, this embodiment decomposes the sensitivity of the temperature peak to the design variables into direct and indirect components. The direct component characterizes the influence of changes in the equivalent thermal conductivity of the design variables on the temperature field and the temperature peak. The indirect component characterizes the influence of changes in the equivalent elastic modulus and structural stiffness of the design variables, causing sequential changes in contact pressure, interfacial thermal resistance, and temperature field, on the temperature peak. The indirect component is calculated using a dual-adjoint method. First, the thermal adjoint equation is solved to obtain the sensitivity of the temperature peak to the interfacial thermal resistance of each contact node. Then, this sensitivity is combined node by node with the sensitivity of the interfacial thermal resistance to the contact pressure to form a comprehensive weight. All node comprehensive weights are merged into an equivalent force vector. The structural adjoint vector is obtained by solving the first-order structural adjoint equation, thereby determining the indirect sensitivity. The direct and indirect components are added to obtain the complete sensitivity, which is used as the gradient information for updating the design variables using MMA.

[0058] Step 106: Update each design variable under constraints based on the sensitivity, and output the converged geometry of the heat dissipation structure. Specifically, using the complete sensitivity obtained in the preceding steps as gradient information, the Moving Asymptote Method (MMA) is used to solve the optimization subproblem. MMA is a mature algorithm for handling large-scale nonlinear constraint optimization. In each iteration, a locally convex approximation is constructed and solved using the current function value and gradient to obtain the updated design variables. The constraint is that the total volume of the heat dissipation structure material does not exceed a preset upper limit, which is determined by the product weight index or available space.

[0059] In one embodiment, after the update, density filtering is applied to each design variable. Specifically, the density filtering operation involves applying density filtering within a filter radius centered on each cell. Within the range, replace the density value of the cell by taking a weighted average of all neighboring cell density values ​​inversely proportional to their distance from the center.

[0060] The specific weighting function for density filtering is: ; in For unit The set of all units within the filtering neighborhood; For unit The center coordinates (in meters). Neighborhood unit The center coordinates (in meters); These are linear weighting coefficients (unit: m), meaning the closer the distance, the greater the weight, and the distance reaches... The time weight is reduced to 0; Neighborhood unit The original design variable values; These are the filtered design variable values.

[0061] in This corresponds to the minimum feature size allowed by the target manufacturing process, such as 1.5mm for aluminum alloy die casting. Density filtering eliminates numerical noise caused by alternating 0-1 density values ​​between adjacent cells, while indirectly controlling the minimum wall thickness to be no less than [a certain value]. .

[0062] Optionally, the constraints also include manufacturing process constraints. For die casting, manufacturing process constraints include the following: applying monotonicity constraints to design variables along the mold opening and closing direction, specifically taking the cumulative maximum value column by column along the draft direction, ensuring that there is material below the position above, satisfying the draft requirements; controlling the minimum wall thickness through the filtering radius of density filtering; and checking whether cells with a density higher than the 0.5 threshold form a connected region after each update, and forcibly setting the design variables of isolated cells not connected to the main body to zero to prevent the formation of closed island structures.

[0063] Repeat the above steps until the convergence condition is met, and output the geometry of the heat dissipation structure. Repeat all steps from contact pressure distribution calculation to design variable update. Each iteration yields a new contact pressure distribution, interface thermal resistance distribution, peak temperature, and sensitivity, and updates the design variables accordingly. The iteration terminates when both of the following conditions are met: the absolute value of the difference between the peak temperatures obtained in two adjacent iterations is less than a preset temperature threshold, and the maximum change of each design variable in two adjacent iterations is less than a preset variable threshold.

[0064] The preset temperature threshold is determined based on the measurement uncertainty of the power device temperature. Half of the measurement uncertainty is taken as the optimization accuracy requirement. For example, when the measurement uncertainty of the MOSFET temperature is ±1.5K, the preset temperature threshold is configured to be 0.5K to 0.8K. The preset variable threshold is, for example, 0.01. A maximum iteration limit is also set as a protection measure, for example, 200 iterations. Since the structural field has a unidirectional influence on the thermal field (geometric changes alter pressure and temperature, but temperature changes do not alter structural deformation), the iterations generally converge within 30 to 80 iterations.

[0065] After the iteration terminates, a threshold is applied to the final density field. The unit is retained as an entity. The elements are deleted to obtain a clearly defined heat dissipation structure geometry, which is then exported as a CAD file for subsequent engineering use.

[0066] In some embodiments, mass production assembly fluctuations are further incorporated into the optimization. Specifically, when applying assembly load in the aforementioned steps, random disturbances within a preset ratio range are applied to the axial preload of each fastening point to generate multiple sets of load conditions. For example, five sets of disturbance values ​​are taken in a uniform distribution within ±15% of the nominal value to simulate the accuracy deviation of the torque gun on the production line. The aforementioned steps are then performed on each condition to obtain multiple sets of temperature peak values.

[0067] The average of multiple temperature peak values and standard deviation According to the preset robustness weighting coefficient Weighted summation is used as the optimization objective instead of the peak temperature under a single operating condition. ; This is the arithmetic mean of multiple temperature peak values, in K. The standard deviation of multiple temperature peak values, in K; The robustness weighting coefficient controls the optimizer's sensitivity to temperature fluctuations. The standard deviation of multiple temperature peaks is determined by gradually increasing the threshold based on whether it exceeds a preset consistency threshold. The specific process is as follows: Initially, take... =0 Run optimization, check if the standard deviation of the peak temperature under multiple operating conditions exceeds the consistency threshold (e.g., 3K); if it does, The optimization is repeated in increments of 0.5 until the standard deviation falls below the consistency threshold. The actual value range is generally between 0.5 and 2.0.

[0068] Taking a typical 3-cell lithium-ion battery pack as an example: the internal dimensions of the casing are, for example, 80mm × 50mm × 20mm. Four MOSFETs are connected in parallel on the protection board, with each MOSFET generating 2.5W of heat at a 5C charging peak. The heat dissipation structure has a design area of ​​60mm × 35mm × 5mm, secured by two M2 screws with a center-to-center distance of 45mm. TIM parameters: =2.5×10 -4 m 2 ·K / W, =50kPa, =0.6, =10kPa.

[0069] Traditional methods use fixed interfacial thermal resistance Optimization resulted in an actual contact pressure of only 15 kPa at the midpoint between the two screws (where the MOSFET is located), corresponding to an actual area thermal resistance of approximately 5.4 × 10⁻⁶ kPa. -4 m 2 The K / W ratio is 2.2 times the design assumption. After optimization, a locally thickened region (increasing from 1.2mm to 2.8mm) is automatically generated directly above the central MOSFET. The increased bending stiffness of this region raises the contact pressure to 42kPa, corresponding to a decrease in the area-to-surface thermal resistance to 2.8×10⁻⁶. -4 m 2 • K / W. Under the same volume constraint, the peak temperature of the MOSFET using this method is reduced by 8 to 12 K; the deviation between simulation prediction and post-assembly measurement is reduced from 15 to 20 K in the traditional method to less than 3 K. The additional material in the thickened region accounts for approximately 12% of the total volume, which is compensated by the thinning of the edge regions away from the heat source.

[0070] Figure 5 This is a schematic diagram comparing the contact thermal performance before and after optimization of the heat dissipation structure in one embodiment of this application. Figure 5 As shown, under the same volume constraint, when traditional optimization is performed using a fixed interface thermal resistance, the contact pressure of the heat dissipation structure in the power device region between the two fastening points is low, resulting in a high actual interface thermal resistance and a high temperature peak in this region. This embodiment considers the coupling relationship between structural stiffness, contact pressure, and interface thermal resistance during the optimization process. It creates a locally thickened structure above the power device, increasing the bending stiffness and contact pressure in this region and correspondingly reducing the interface thermal resistance. Simultaneously, it compensates for the increased material due to the local thickening by appropriately thinning the area away from the heat source. Therefore, while satisfying the overall volume constraint, it improves the contact thermal conductivity of the power device region and reduces the temperature peak during the charging cycle.

[0071] Therefore, this method introduces structural mechanics analysis to predict the contact pressure distribution in each iteration of the heat dissipation structure topology optimization. The pressure-thermal resistance characteristics of the thermal interface material are converted into a position-dependent interface thermal resistance distribution and then fed back into the thermal analysis. The sensitivity calculation simultaneously includes the components of the design variables that directly affect the temperature field through changes in material thermal conductivity and the components that indirectly affect the temperature field through changes in structural stiffness. This allows the optimizer to autonomously find a balance between thermal conductivity and contact pressure uniformity. The resulting heat dissipation structure geometry has a small deviation in predicting the thermal performance under mass production assembly conditions, improves product temperature consistency, and reduces the number of engineering design iterations.

[0072] Please see Figure 6 , Figure 6 This is a schematic block diagram of an intelligent design device for a battery pack heat dissipation structure according to this application. Figure 6 As shown, the device includes: Design domain construction module 601 is used to determine the design domain of the heat dissipation structure according to the internal component layout of the battery pack, divide the design domain into a grid and assign design variables to each unit. The contact pressure calculation module 602 is used to apply assembly load to the structure corresponding to the current design variables and solve the structural equilibrium equation according to the heat dissipation structure assembly scheme, so as to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. The interface thermal resistance mapping module 603 is used to map the contact pressure distribution to the interface thermal resistance distribution at each position of the contact surface based on the pressure-thermal resistance characteristics of the thermally conductive interface material. Thermal analysis module 604 is used to perform thermal analysis on the heat dissipation structure with the interface thermal resistance distribution as the boundary condition to obtain the temperature peak of the power device. The sensitivity calculation module 605 is used to calculate the sensitivity of the temperature peak to each design variable. The sensitivity includes a direct component and an indirect component. The direct component corresponds to the influence of the design variable on the temperature field caused by the change in the thermal conductivity of the material. The indirect component corresponds to the influence of the design variable on the temperature field caused by the change in the contact pressure distribution and the interface thermal resistance distribution due to the change in the structural stiffness. The design variable update and iteration control module 606 is used to update each design variable under constraints according to the sensitivity; repeat the above steps of solving the structural equilibrium equation to update each design variable until the convergence condition is met, and output the converged heat dissipation structure geometry.

[0073] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0074] Figure 7 This is a schematic diagram of the structure of an electronic device according to one embodiment of this application. Figure 7As shown, at the hardware level, the electronic device includes a processor 1102, an internal bus 1104, a network interface 1106, memory 1108, and non-volatile memory 1110, and may also include other hardware required for business operations. One or more embodiments of this specification can be implemented in software, for example, the processor 1102 reads the corresponding computer program from the non-volatile memory 1110 into the memory 1108 and then runs it. Of course, in addition to software implementation, one or more embodiments of this specification do not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution subject of the following processing flow is not limited to each logic module, but can also be hardware or logic devices.

[0075] Based on the same concept as the above-described method, this application also provides a machine-readable storage medium storing a plurality of computer instructions. When executed by a processor, these computer instructions can implement the intelligent design method for the battery pack heat dissipation structure disclosed in the above examples of this application. The machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device, and can contain or store information such as executable instructions, data, etc.

[0076] Based on the same application concept as the above method, this application embodiment also provides a computer program product, the computer program product including a computer program, which, when executed by a processor, implements the intelligent design method for battery pack heat dissipation structure disclosed in the above examples of this application.

[0077] In the above embodiments, the descriptions of each embodiment have different focuses. Parts not described in detail in a certain embodiment can be referred to in the relevant descriptions of other embodiments. The above descriptions are merely preferred embodiments of this application and explanations of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by specific combinations of the above technical features, but should also cover other technical solutions formed by arbitrary combinations of the above technical features or their equivalent features without departing from the inventive concept.

Claims

1. A smart design method for a battery pack heat dissipation structure, characterized in that, include: The design domain of the heat dissipation structure is determined based on the internal component layout of the battery pack. The design domain is then divided into a grid, and design variables are assigned to each unit. Based on the heat dissipation structure assembly scheme, the assembly load is applied to the structure corresponding to the current design variables and the structural equilibrium equation is solved to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. Based on the pressure-thermal resistance characteristics of the thermally conductive interface material, the contact pressure distribution is mapped to the interface thermal resistance distribution at each position of the contact surface. Thermal analysis was performed on the heat dissipation structure using the interface thermal resistance distribution as a boundary condition to obtain the peak temperature of the power device. Calculate the sensitivity of the peak temperature to each design variable. The sensitivity includes a direct component and an indirect component. The direct component corresponds to the influence of the design variable on the temperature field through the change in the thermal conductivity of the material. The indirect component corresponds to the influence of the design variable on the temperature field through the change in the contact pressure distribution and the interface thermal resistance distribution caused by the change in the structural stiffness. Update each design variable under constraints based on the sensitivity; repeat the above steps of solving the structural equilibrium equations and updating each design variable until the convergence condition is met, and output the converged heat dissipation structure geometry.

2. The method according to claim 1, characterized in that, The process of applying assembly loads to the structure corresponding to the current design variables and solving the structural equilibrium equations to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device includes: calculating the axial preload at each fastening point based on the preload torque and fastener specification parameters of each fastener in the heat dissipation structure; setting up normal contact pairs between the heat dissipation structure and the power device; assembling the stiffness matrix using the axial preload as the load and the equivalent elastic modulus corresponding to each unit design variable and solving the displacement field; and extracting the normal contact pressure at each node of the contact surface from the displacement field to form the contact pressure distribution.

3. The method according to claim 1, characterized in that, The step of mapping the contact pressure distribution to the interfacial thermal resistance distribution at each location on the contact surface includes: fitting the measured thermal resistance data of the thermally conductive interface material under different pressures to obtain a monotonically decreasing mapping function with pressure as the independent variable and area specific thermal resistance as the dependent variable; substituting the contact pressure value at each location on the contact surface into the monotonically decreasing mapping function to calculate the area specific thermal resistance value at each location; and setting the area specific thermal resistance of locations with contact pressure lower than a preset minimum effective pressure threshold as a cutoff upper limit value in a smooth transition manner.

4. The method according to claim 3, characterized in that, The monotonically decreasing mapping function adopts a power law form, with the nominal thermal resistance value under the reference pressure and the reference pressure as the reference parameters, and the negative power of the ratio of the local contact pressure to the reference pressure as the scaling factor. The power exponent is determined by least-squares fitting of the measured thermal resistance data of the thermally conductive interface material at multiple pressure points.

5. The method according to claim 1, characterized in that, The thermal analysis is a transient thermal analysis covering the entire charging cycle; the heat source load is calculated based on the charging current curve and the on-resistance of the power device at the operating temperature; the temperature peak is the highest temperature reached by the power device during the entire charging cycle; the time of occurrence of the temperature peak and the corresponding temperature field distribution are recorded simultaneously with obtaining the temperature peak.

6. The method according to claim 3, characterized in that, The calculation of the indirect components includes: solving the thermal adjoint equation based on the temperature field distribution at the moment the temperature peak occurs to obtain the thermal adjoint vector; calculating the sensitivity of the temperature peak to the interfacial thermal resistance of each node on the contact surface based on the thermal adjoint vector; multiplying the sensitivity of the temperature peak to the interfacial thermal resistance of each node by the sensitivity of the interfacial thermal resistance at each node to the contact pressure node by node to obtain the comprehensive weight of each node; combining the comprehensive weights of each node to form an equivalent force vector, which is used as the right-hand side term of the structural adjoint equation to obtain the structural adjoint vector; and calculating the indirect components of each design variable based on the structural adjoint vector and the partial derivatives of each element stiffness matrix with respect to the design variables.

7. The method according to claim 6, characterized in that, The sensitivity of the interfacial thermal resistance at each node to the contact pressure is obtained by differentiating the monotonically decreasing mapping function with respect to pressure, and its value is negative; the equivalent force vector combines the comprehensive weights of all nodes on the contact surface into a single right-hand term, compressing the original multiple structural adjoint equations that needed to be solved separately for each contact node into a single solution.

8. The method according to claim 1, characterized in that, The equivalent elastic modulus and equivalent thermal conductivity of each unit are determined by the penalty interpolation method for solid isotropic materials, which is obtained by multiplying the penalty power of the design variables of each unit with the corresponding property value of the matrix material; the penalty power is a preset constant with a value range of 2 to 4.

9. A smart design device for a battery pack heat dissipation structure, characterized in that, include: The design domain construction module is used to determine the design domain of the heat dissipation structure based on the layout of the internal components of the battery pack, divide the design domain into a grid and assign design variables to each unit. The contact pressure calculation module is used to apply assembly loads to the structure corresponding to the current design variables and solve the structural equilibrium equations according to the heat dissipation structure assembly scheme, so as to obtain the contact pressure distribution on the contact surface between the heat dissipation structure and the power device. The interface thermal resistance mapping module is used to map the contact pressure distribution to the interface thermal resistance distribution at each position of the contact surface based on the pressure-thermal resistance characteristics of the thermally conductive interface material. The thermal analysis module is used to perform thermal analysis on the heat dissipation structure with the interface thermal resistance distribution as the boundary condition to obtain the temperature peak of the power device. The sensitivity calculation module is used to calculate the sensitivity of the temperature peak to each design variable. The sensitivity includes a direct component and an indirect component. The direct component corresponds to the influence of the design variable on the temperature field caused by the change in the thermal conductivity of the material. The indirect component corresponds to the influence of the design variable on the temperature field caused by the change in the contact pressure distribution and the interface thermal resistance distribution due to the change in the structural stiffness. The design variable update and iteration control module is used to update each design variable under constraints according to the sensitivity; repeat the above steps of solving the structural equilibrium equation to update each design variable until the convergence condition is met, and output the converged heat dissipation structure geometry.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for performing the method as described in any one of claims 1 to 8.

Images