A method and device for constructing a liquid-cooled radiator microchannel with integrated bypass control

Abstract

The present application discloses a method and device for constructing a liquid-cooled radiator microchannel with integrated bypass control, which is used to solve the energy waste problem caused by fixed high-resistance microchannels. The method of the present application includes: establishing a three-dimensional geometric model of the liquid-cooled radiator, the three-dimensional geometric model including a main channel for processing peak heat load conditions, a bypass channel for processing low heat load conditions, and an SMA microactuator arranged in the bypass channel; defining the fluid domain, solid domain, and actuator domain respectively, and configuring an SMA constitutive model for the actuator domain to describe the relationship between temperature, stress, and phase change; establishing a bidirectional coupling relationship between the fluid domain, solid domain, and actuator domain based on the SMA constitutive model to form a digital twin model; applying a simulated heat load that changes with time to the digital twin model to obtain the simulation performance parameters of the liquid-cooled radiator under dynamic conditions; iteratively optimizing the target design parameters based on the simulation performance parameters to obtain the target design parameters.

Description

Technical Field

[0001] The present application relates to the field of heat dissipation control technology, and in particular to a method and device for constructing a liquid-cooled radiator microchannel with integrated bypass control. Background Art

[0002] With the rapid development of semiconductor technology, the power density and heat flux of electronic components such as central processing units (CPUs), graphics processing units (GPUs), 5G radio frequency devices, and power semiconductors have skyrocketed. Traditional air cooling technology is no longer able to meet the heat dissipation requirements of these high-heat-flux scenarios. Therefore, liquid cooling technology, particularly microchannel liquid cooling, has become a core development direction for addressing high-heat-flux dissipation due to its efficient heat transfer capabilities and compact structure.

[0003] In existing technology, the core structure of a microchannel liquid-cooled heat sink is typically a large array of high-aspect-ratio parallel microchannels with widths ranging from tens to hundreds of microns, fabricated on a highly thermally conductive substrate such as copper or silicon using precision etching or machining methods. This design greatly increases the contact surface area between the coolant and the heat dissipation substrate, resulting in an extremely high convective heat transfer coefficient. However, in order to prevent electronic components from overheating under any circumstances, the heat sink structure of mainstream microchannel liquid-cooled heat sinks is usually statically designed to handle peak heat loads.

[0004] However, existing microchannel structures only have a fixed microchannel path and exhibit extremely high fluid resistance. Microchannel structures enhance heat transfer by significantly increasing the surface area to volume ratio. However, the large surface area and tiny channel size amplify the viscous friction effect, requiring extremely high pumping power to overcome the enormous flow resistance. Even when the chip is in a low-power state, the cooling system still consumes high pumping power to drive the coolant through these high-resistance channels to maintain circulation, resulting in significant energy waste. Summary of the Invention

[0005] The present application provides a method and device for constructing a liquid-cooled radiator microchannel with integrated bypass control, which is used to solve the energy waste problem caused by fixed high-flow resistance microchannels.

[0006] A first aspect of the present application provides a method for constructing a liquid-cooled radiator microchannel with integrated bypass control, comprising:

[0007] Establishing a three-dimensional geometric model of a liquid-cooled radiator, the three-dimensional geometric model including a main channel for handling peak heat load conditions, a bypass channel for handling low heat load conditions, and an SMA microactuator disposed in the bypass channel;

[0008] Performing mesh generation processing for finite element numerical simulation on the three-dimensional geometric model to obtain a mesh model;

[0009] In the grid model, corresponding to the cooling fluid of the liquid-cooled radiator, the heat dissipation substrate, and the SMA microactuator, a fluid domain, a solid domain, and an actuator domain are defined respectively, and an SMA constitutive model for describing the relationship between temperature, stress, and phase change is configured for the actuator domain;

[0010] establishing a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model to form a digital twin model, wherein the bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling;

[0011] Applying a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions;

[0012] The target design parameters are iteratively optimized according to the simulation performance parameters to obtain the target design parameters, wherein the target design parameters include at least one of the geometric size, shape, anchor point position and material phase change activation temperature of the SMA microactuator.

[0013] Optionally, establishing a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model includes:

[0014] performing a bidirectional heat transfer calculation among all contact surfaces of the fluid domain, the solid domain, and the actuator domain to obtain a temperature field and a heat flux density distribution to establish a conjugate heat transfer coupling;

[0015] Applying the temperature field as a thermal load to the actuator domain, calculating the phase change and the resulting phase change strain caused by the temperature change through the SMA constitutive model, and solving to obtain the macroscopic deformation displacement of the SMA microactuator to establish thermal-mechanical deformation coupling;

[0016] The geometric boundaries of the fluid domain are updated in real time according to the macroscopic deformation displacement. At the same time, the pressure and shear force generated when the fluid domain flows through the surface of the SMA microactuator are fed back as mechanical loads to the structural mechanics calculation of the actuator domain to establish fluid-solid coupling.

[0017] Optionally, performing meshing processing on the three-dimensional geometric model for finite element numerical simulation to obtain a mesh model includes:

[0018] Generate multi-layer refined boundary layer meshes at the interface between the fluid region and all solid walls;

[0019] Meshing the area away from the SMA microactuator using a static grid;

[0020] The SMA microactuator region and the fluid region adjacent to the SMA microactuator are meshed using a dynamic deformation grid or an overlapping grid.

[0021] Optionally, configuring an SMA constitutive model for the actuator domain to describe the relationship between temperature, stress and phase change includes:

[0022] Obtaining phase change parameters of the SMA microactuator, wherein the material characteristic parameters include at least Young's modulus, phase change stress influence coefficient, and theoretical phase change temperature;

[0023] Performing a differential scanning calorimetry experiment on a sample made from the same batch of materials as the SMA microactuator to determine the precise phase transition temperature of the sample under actual thermal cycling, and correcting the theoretical phase transition temperature in the phase transition parameters based on the precise phase transition temperature, and establishing an SMA constitutive model based on the corrected phase transition parameters;

[0024] After applying a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions, the method further includes:

[0025] The simulated performance parameters are compared with the benchmark performance parameters measured through physical prototype experiments to obtain performance deviations, and an inverse correction algorithm is used to iteratively adjust one or more phase change parameters in the SMA constitutive model until the performance deviation meets the preset requirements.

[0026] Optionally, the model architecture of the SMA constitutive model is a Brinson architecture or an Auricchio architecture.

[0027] Optionally, the liquid cooling radiator microchannel construction method further includes:

[0028] Configuring a damage evolution module for the SMA constitutive model, wherein the damage evolution module is used to dynamically modify the phase change parameters according to the number of thermodynamic cycles;

[0029] During the simulation process, the SMA constitutive model is continuously updated according to the damage evolution module, and the performance indicators of the SMA microactuator are monitored;

[0030] Verify whether the attenuation of the performance indicator meets the design redundancy range after reaching the preset cycle life number.

[0031] Optionally, iteratively optimizing target design parameters according to the simulation performance parameters to obtain target design parameters includes:

[0032] Iteratively optimizing the target design parameters according to the simulation performance parameters and the preset optimization targets to obtain the target design parameters;

[0033] The preset optimization objectives include at least minimizing the fluid pumping power of the liquid-cooled radiator under low heat load conditions and minimizing the response time of the SMA microactuator from open to closed or from closed to open.

[0034] A second aspect of the present application provides a liquid-cooled radiator microchannel construction device with integrated bypass control, comprising:

[0035] a first modeling unit, configured to establish a three-dimensional geometric model of the liquid-cooled radiator, the three-dimensional geometric model comprising a main channel for processing a peak heat load condition, a bypass channel for processing a low heat load condition, and an SMA microactuator disposed in the bypass channel;

[0036] A grid unit is used to perform a grid division process on the three-dimensional geometric model for finite element numerical simulation to obtain a grid model;

[0037] defining a unit for defining a fluid domain, a solid domain, and an actuator domain in the grid model corresponding to the cooling fluid, the heat dissipation substrate, and the SMA microactuator of the liquid-cooled radiator, respectively, and configuring an SMA constitutive model for describing the relationship between temperature, stress, and phase change for the actuator domain;

[0038] a second modeling unit, configured to establish a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model to form a digital twin model, wherein the bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling;

[0039] A simulation unit, configured to apply a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions;

[0040] An optimization unit is configured to iteratively optimize target design parameters according to the simulation performance parameters to obtain target design parameters, wherein the target design parameters include at least one of the geometric size, shape, anchor point position, and material phase change activation temperature of the SMA microactuator.

[0041] A third aspect of the present application provides a liquid-cooled radiator microchannel construction system with integrated bypass control, the system comprising:

[0042] processor, memory, input and output units, and buses;

[0043] The processor is connected to the memory, the input and output unit, and the bus;

[0044] The memory stores a program, and the processor calls the program to execute the first aspect and any optional method for constructing a liquid-cooled radiator microchannel with integrated bypass control in the first aspect.

[0045] The fourth aspect of the present application provides a computer-readable storage medium having a program stored thereon, which, when executed on a computer, executes the first aspect and any optional method for constructing a liquid-cooled radiator microchannel with integrated bypass control in the first aspect.

[0046] It can be seen from the above technical solutions that this application has the following advantages:

[0047] Two parallel fluid paths, a main channel and a bypass channel, were constructed using a three-dimensional geometric model. The main channel handles peak heat loads, while an SMA microactuator is installed in the bypass channel, providing a physical mechanism for switching fluid paths. The fluid, solid, and actuator domains were defined for the three-dimensional geometric model, and an SMA constitutive model was configured for the actuator domain. By establishing conjugate heat transfer coupling, thermal-mechanical-deformation coupling, and fluid-solid coupling relationships between the fluid, solid, and actuator domains, a dynamic simulation digital twin model was generated. This digital twin model can fully simulate the radiator's operating process: changes in heat load cause changes in the temperature field, which in turn drive the microactuator's movement. The microactuator's movement changes the fluid path, which in turn affects the system's pressure drop and temperature field.

[0048] By using this digital twin model for iterative optimization and applying a varying simulated thermal load, the digital twin model can output simulation performance parameters under different working conditions. Based on these simulation performance parameters, the design parameters such as the geometric dimensions, anchor point position, or phase change activation temperature of the SMA microactuator can be optimized and adjusted. This process enables the construction of a liquid-cooled radiator microchannel with integrated bypass control, solving the energy waste problem caused by fixed high-flow resistance structures. That is, it ensures that under low thermal load conditions, the SMA microactuator can accurately keep the bypass channel open, allowing the coolant to flow through the low-resistance path, thereby reducing the total flow resistance of the system and directly reducing the consumption of pumping power. Under high thermal loads, the actuator closes the bypass to ensure that all the coolant flows through the main channel to meet the heat dissipation requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in this application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1A schematic flow chart of an embodiment of a method for constructing a liquid-cooled radiator microchannel with integrated bypass control provided in this application;

[0051] Figure 2 A schematic flow chart of another embodiment of the method for constructing a liquid-cooled radiator microchannel with integrated bypass control provided by the present application;

[0052] Figure 3 This is a schematic structural diagram of an embodiment of a liquid-cooled radiator microchannel construction device with integrated bypass control provided by the present application;

[0053] Figure 4 This is a schematic structural diagram of an embodiment of a liquid-cooled radiator microchannel construction system with integrated bypass control provided in this application. DETAILED DESCRIPTION

[0054] The present application provides a method and device for constructing a liquid-cooled radiator microchannel with integrated bypass control, which is used to solve the energy waste problem caused by fixed high-flow resistance microchannels.

[0055] See also Figure 1 , Figure 1 An embodiment of a method for constructing a liquid-cooled radiator microchannel with integrated bypass control provided in this application includes:

[0056] 101. Establish a three-dimensional geometric model of a liquid cooling radiator, the three-dimensional geometric model including a main channel for handling peak heat load conditions, a bypass channel for handling low heat load conditions, and an SMA microactuator disposed in the bypass channel;

[0057] A 3D geometric model of the liquid-cooled heat sink was constructed using modeling software, digitally describing its shape, structure, and dimensions in 3D computer space. This 3D geometric model includes three core components: the main channel, the bypass channel, and the SMA microactuator. By connecting the main and bypass channels in parallel, two alternative paths are provided for the coolant: a high-resistance, high-heat-transfer path and a low-resistance, low-heat-transfer path. Furthermore, the SMA microactuator provides a physical entity for switching between these two paths. An SMA microactuator is a solid component model with a specific geometric shape, specifically made of shape memory alloy (SMA). In the 3D geometric model, the SMA microactuator can be modeled as a separate part, such as a cantilever beam or thin sheet with a specific thickness, length, and width. The size and position of the SMA microactuator in the 3D geometric model determine whether deformation can effectively alter the flow state of the bypass channel.

[0058] Ideally, an SMA microactuator requires no external power or control and responds precisely to temperature changes. When the device is under low load and the coolant temperature is below its phase change activation point, the SMA microactuator remains stable in a preset low-temperature configuration (e.g., a bent cantilever beam), fully opening the low-resistance bypass channel and minimizing pumping power consumption. Once the device load increases and the coolant temperature rises to its phase change activation point, the SMA microactuator undergoes a precise phase change, driving itself back to its memorized high-temperature configuration (e.g., a straight beam), thereby closing or completely sealing the bypass channel entrance and forcing all coolant into the main microchannel to handle peak heat dissipation. When the heat load disappears and the liquid temperature drops, the SMA microactuator automatically resets to its low-temperature configuration with the bypass open, driven by fluid dynamics, thereby completing the intelligent allocation of the fluid path.

[0059] It should be noted that during design, it is necessary to ensure that the flow resistance of the bypass channel in the open state is much lower than the total resistance of the main channel. It can adopt a larger diameter than the main channel and keep its shape smooth and without sharp turns to minimize pressure loss caused by viscous friction and shape changes. A valve seat structure and an anchor point structure for mounting and fixing the SMA microactuator need to be designed on the channel wall in the closed position of the SMA microactuator in the bypass channel. When the SMA microactuator is deformed by heat, it can fit tightly with the valve seat to effectively cut off or seal the bypass channel and prevent coolant leakage. The local width and height of the bypass channel must provide sufficient, interference-free space for the complete movement of the SMA microactuator, that is, from the cold open position to the hot closed position.

[0060] 102. Performing meshing processing on the three-dimensional geometric model for finite element numerical simulation to obtain a mesh model;

[0061] Since computers cannot directly solve partial differential physics equations defined in continuous space, the continuous geometric space must first be divided into a limited number of tiny units with simple shapes. This division process is called meshing. In order to perform finite element numerical simulation on a three-dimensional geometric model, the three-dimensional geometric model needs to be meshed and converted into a discretized mesh model. Through this process, the original continuous solution domain is replaced by a set of nodes and units, and the physical equations will be converted into a set of algebraic equations on these discrete units and nodes, so that the computer can perform numerical solutions. The specific implementation can be completed through the meshing module in the simulation software. The software reads the three-dimensional geometric model and automatically or semi-automatically generates a mesh within the entire volume of the model based on preset rules and parameters.

[0062] 103. In the mesh model, corresponding to the cooling fluid, heat dissipation substrate and SMA microactuator of the liquid cooling radiator, the fluid domain, solid domain and actuator domain are defined respectively, and the SMA constitutive model used to describe the relationship between temperature, stress and phase change is configured for the actuator domain;

[0063] The mesh model generated in step 102 is only a geometrically discrete model and does not contain any physical properties. Before simulation, it is necessary to select the mesh unit sets that constitute different components to define their physical property domains. First, the mesh set corresponding to the space occupied by the coolant is defined as the fluid domain, and fluid mechanics and heat transfer equations, such as the Navier-Stokes equations and the energy equation, are assigned to it. At the same time, the material properties of the coolant, such as density, viscosity, and heat capacity, are input. Secondly, the mesh set corresponding to the heat dissipation substrate is defined as the solid domain, and solid heat transfer equations, such as Fourier's law, are assigned to it. Parameters such as the thermal conductivity and density of the substrate material are also input. For the SMA microactuator, considering that it is both a solid and its internal temperature conduction and structural response under stress need to be calculated, and it is also a smart material, its behavior cannot be described by conventional elastic mechanics models. Therefore, a special actuator domain is defined for the SMA microactuator, and an SMA constitutive model is configured for this actuator domain to describe the relationship between temperature, stress, and phase change.

[0064] The unique shape memory effect of SMA materials stems from the reversible transformation of the internal crystal structure between the two solid phases of martensite and austenite under temperature or stress. Conventional constitutive models cannot describe this phenomenon. Therefore, a specialized SMA constitutive model is required. The core function of this SMA constitutive model is to establish a quantitative relationship that describes the intrinsic connection between the material's temperature, stress, and internal phase transition state. When the temperature or stress in the actuator domain changes, the internal phase transition state can be calculated by calling the SMA constitutive model. This allows the changes in physical properties such as material stiffness and thermal expansion coefficient to be determined, as well as the driving force or deformation caused by the phase transition.

[0065] 104. Based on the SMA constitutive model, a bidirectional coupling relationship is established between the fluid domain, solid domain and actuator domain to form a digital twin model. The bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling and fluid-solid coupling;

[0066] After assigning independent physical properties to the different domains of the mesh model, these independent physical fields need to be connected and rules for their interaction established. In the actual operation of a liquid-cooled heat sink, heat, forces, and fluids interact with each other. Therefore, it is necessary to establish a bidirectional coupling relationship between the fluid domain, solid domain, and actuator domain to simulate the interaction of these physical phenomena, thereby integrating the separate physical models into a digital twin model that reflects the behavior of a real-world liquid-cooled heat sink. In this digital twin model, changes in any physical quantity will trigger corresponding changes in other physical quantities through defined coupling paths, thereby simulating the operation of the SMA microactuator regulating the fluid path.

[0067] Specifically, bidirectional coupling relationships include conjugate heat transfer coupling, thermo-deformation coupling, and fluid-structure coupling. Conjugate heat transfer coupling refers to the heat transfer relationship between the solid and fluid domains, thermo-deformation coupling refers to the causal relationship between the temperature within the actuator domain and its mechanical deformation, and fluid-structure coupling refers to the bidirectional influence between fluid flow and the deformation of the actuator structure.

[0068] 105. Apply a time-varying simulated heat load to the digital twin model to obtain the simulation performance parameters of the liquid cooling radiator under dynamic conditions;

[0069] In order to obtain the performance of the liquid radiator in a scenario close to the actual use, the dynamic working conditions can be reproduced by applying a simulated heat load that changes with time. Specifically, a transient solver is selected to calculate the state of the model at each discrete time point, so as to capture the entire process of the working condition evolving over time. A heat source boundary condition is applied to the area in the model that represents the heat generation of the device (such as a specific surface of the radiator substrate). The power or heat flux density of the heat source can be defined as a function of time, such as a step or ramp function that simulates the device starting from standby state, entering full load operation, and then returning to standby state. At the same time, a constant coolant flow rate and temperature are set for the fluid inlet of the model, and a pressure condition is set for the outlet.

[0070] After running the simulation, the solver will step by step solve the coupled physical equations at each time point within the set total time length, and obtain a series of simulation performance parameters. These simulation performance parameters are indicators for evaluating the performance of the heat sink, mainly including:

[0071] Temperature response: A curve showing the maximum or average temperature of the interface between the device and the liquid heat sink over time, used to evaluate heat dissipation performance and temperature control stability.

[0072] Pressure Drop and Pumping Power: This data shows the pressure drop between the inlet and outlet of a liquid heat sink over time. Pumping power can be calculated based on the pressure drop and inlet flow rate. Pumping power is a key indicator of system energy consumption.

[0073] Flow distribution: The change of coolant flow through the main channel and bypass channel over time is used to directly observe the flow distribution effect of the SMA microactuator;

[0074] Actuator displacement: The change in the deformation displacement of the SMA microactuator over time is used to confirm the response speed and stroke of its action.

[0075] 106. Iteratively optimize the target design parameters according to the simulation performance parameters to obtain the target design parameters, where the target design parameters include at least one of the geometric size, shape, anchor point position and material phase change activation temperature of the SMA microactuator.

[0076] The simulation results only reflect the performance of a specific design solution. In order to find the optimal design that enables the liquid heat sink to achieve the preset performance target, it is necessary to iteratively optimize the target design parameters using genetic algorithms or particle swarm optimization algorithms based on the simulation performance parameters. The target design parameters specifically refer to the parameters that are allowed to change during the optimization process. The target design parameters include at least one of the geometric size, shape, anchor point location, and material phase change activation temperature of the SMA microactuator. Specifically:

[0077] Geometric size and shape: determine the stiffness, thermal response speed and flow channel blocking effect of the SMA microactuator;

[0078] Anchor point position: changes the motion trajectory and force arm of the SMA microactuator during deformation, affecting its efficiency and sealing performance in closing the bypass channel;

[0079] Material phase change activation temperature: This directly determines the temperature threshold at which the SMA microactuator operates. The optimization of this parameter is intended to precisely match the heat sink's adaptive behavior with the optimal operating temperature range of the cooled chip.

[0080] The optimization process is to find the optimal solution under multiple objectives. For example, an SMA microactuator with extremely fast response speed can be thinner, which may lead to insufficient structural strength or poor sealing effect. The purpose of optimization is to find the best balance between these contradictory performance indicators. It should be noted that this application mainly focuses on how to design and arrange the SMA microactuators in the microchannels, while the microchannels of the main channel and bypass channel themselves can be constructed through existing generation algorithms. For example, the topology optimization algorithm, under a given design space and boundary conditions, takes maximizing heat dissipation efficiency or minimizing flow resistance as the optimization goal, and can automatically generate non-traditional, efficient flow channel layouts similar to biological fractal structures (such as leaf veins or vascular networks). In addition, generation algorithms based on fractal theory can also be directly used. These algorithms are specifically used to design microchannels that can achieve optimal fluid distribution and heat transfer.

[0081] After the microchannels are constructed, precision techniques such as femtosecond laser micromachining or photolithography are used to fabricate the main channel, bypass channel, and SMA microactuator anchoring structure on a highly thermally conductive substrate. Furthermore, based on the target design parameters, SMA microactuators with optimized shape and size are fabricated through processes such as laser microcutting. The prefabricated SMA microactuators are then precisely placed onto the substrate's anchoring points and permanently secured using laser micro-spot welding or highly thermally conductive epoxy adhesive. Finally, the cover plate and the fabricated substrate are integrally bonded and packaged using methods such as diffusion welding or brazing to form a closed fluid circuit.

[0082] In this embodiment, two parallel fluid paths, a main channel and a bypass channel, are constructed using a three-dimensional geometric model. The main channel is used to handle peak heat loads, while an SMA microactuator is installed in the bypass channel, providing a physical mechanism for switching fluid paths. The fluid domain, solid domain, and actuator domain are defined for the three-dimensional geometric model, and an SMA constitutive model is configured for the actuator domain. By establishing conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling relationships between the fluid domain, solid domain, and actuator domain, a dynamic simulation digital twin model is obtained. This digital twin model can fully simulate the working process of the radiator: changes in heat load cause changes in the temperature field, which drive the microactuator to operate. The microactuator's operation changes the fluid path, and the change in the fluid path in turn affects the system pressure drop and temperature field.

[0083] By using this digital twin model for iterative optimization and applying a varying simulated thermal load, the digital twin model can output simulation performance parameters under different working conditions. Based on these simulation performance parameters, the design parameters such as the geometric dimensions, anchor point position, or phase change activation temperature of the SMA microactuator can be optimized and adjusted. This process enables the construction of a liquid-cooled radiator microchannel with integrated bypass control, solving the energy waste problem caused by fixed high-flow resistance structures. That is, it ensures that under low thermal load conditions, the SMA microactuator can accurately keep the bypass channel open, allowing the coolant to flow through the low-resistance path, thereby reducing the total flow resistance of the system and directly reducing the consumption of pumping power. Under high thermal loads, the actuator closes the bypass to ensure that all the coolant flows through the main channel to meet the heat dissipation requirements.

[0084] The following is a detailed description of the method for constructing a liquid cooling radiator microchannel with integrated bypass control provided by this application. Figure 2 , Figure 2 Another embodiment of the method for constructing a liquid-cooled radiator microchannel with integrated bypass control provided by the present application includes:

[0085] 201. Establish a three-dimensional geometric model of a liquid cooling radiator, the three-dimensional geometric model including a main channel for handling peak heat load conditions, a bypass channel for handling low heat load conditions, and an SMA microactuator disposed in the bypass channel;

[0086] In this embodiment, step 201 is similar to step 101 in the aforementioned embodiment and will not be described again here.

[0087] 202. Generate multi-layer encrypted boundary layer grids in the interface area between the fluid region and all solid walls;

[0088] When a fluid flows past a solid wall, its velocity is zero at the wall due to viscous effects and then changes dramatically to the mainstream velocity within a very thin region, the boundary layer. Simultaneously, heat exchange between the wall and the fluid occurs primarily within this thin layer, resulting in sharp temperature gradients. Using a conventional, larger mesh size would fail to capture these dramatic gradients, leading to significant errors in the calculated wall shear force, thus the flow resistance, and the convective heat transfer coefficient, thus compromising the reliability of the overall simulation results. Therefore, a multi-layered, dense boundary layer mesh is generated for the interface between the fluid region and all solid walls in the 3D geometry model, including the walls of the main and bypass channels, as well as the surface of the SMA microactuator. This mesh structure consists of multiple layers of thin mesh elements arranged parallel to and closely adjacent to the solid wall. The boundary layer mesh parameters primarily include the height of the first layer, the total number of layers, and the thickness growth rate between layers. The first layer height is the most critical parameter, as its value directly impacts the ability to resolve near-wall flows. In engineering calculations, it can be determined by a dimensionless parameter. The specific value needs to be matched according to the selected turbulence model to ensure that the calculation of velocity and temperature falls within the effective range of the model.

[0089] 203. Use static mesh to mesh the area far away from the SMA microactuator;

[0090] In the entire three-dimensional geometric model, only the geometric boundaries of the SMA microactuator and its adjacent areas will change over time. For those areas whose geometric shapes remain unchanged throughout the simulation process, such as the main channel, most bypass channels, inlet and outlet cavities, etc., there is no need to update their meshes at each calculation time step. Therefore, for areas far away from the SMA microactuator, using a one-time generated and permanently fixed static mesh in these areas can avoid unnecessary mesh reconstruction or node position calculations, thereby significantly shortening the total time required for simulation. Specifically, standard mesh generation algorithms, such as the Delaunay algorithm, can be used to fill conventional mesh units such as tetrahedrons or hexahedrons, and after generation, the position coordinates of these mesh units and their nodes will remain unchanged in subsequent transient calculations.

[0091] Although it is a static mesh, its density still needs to be controlled according to physical phenomena. For example, in areas with large flow field changes, such as channel inlets or corners, the mesh density should be increased appropriately, while in straight pipe sections with stable flow fields, the mesh density can be appropriately reduced.

[0092] 204. Meshing the SMA microactuator region and the fluid region adjacent to the SMA microactuator using a dynamically deforming grid or an overlaid grid;

[0093] Since the SMA microactuator will deform under temperature drive, its physical boundary will move, thereby changing the geometry of the fluid domain. Static grids cannot cope with such changes in geometric boundaries. Therefore, in this embodiment, dynamic grid technology is used for the SMA microactuator area and the fluid area adjacent to the SMA microactuator to ensure that the computational grid can adapt to the changing geometric boundaries at any time, so that the simulation can correctly calculate the flow field changes caused by the movement of the SMA microactuator. Specifically, the area containing the SMA microactuator and the fluid around it that will be affected is defined as a dynamic area, and the following dynamic grid strategy is adopted:

[0094] Dynamically deformable mesh: The mesh nodes on the surface of the SMA microactuator are set to move with the solid boundary. The built-in smoothing algorithm automatically adjusts the position of the mesh nodes inside the adjacent fluid domain to adapt to the boundary movement, just like stretching or compressing a spring network.

[0095] Overlapping meshes: Create at least two independent meshes: one static background mesh covering the entire fluid domain, and the other component mesh that wraps around the SMA microactuator and moves with it. During calculations, the two meshes exchange data in the overlapping areas using an interpolation algorithm to simulate motion.

[0096] It should be noted that for the bending deformation of the SMA microactuator in this solution, which is similar to a cantilever beam, if the deformation is small, a dynamically deforming mesh with higher computational efficiency can be selected. For complex motion, an overlapping mesh with better mesh robustness can be selected. When using a dynamically deforming mesh, if excessive deformation can cause mesh element distortion and quality degradation, an automatic mesh reconstruction mechanism should be enabled. This mechanism regenerates the mesh in poor quality areas during computational interruptions to ensure computational stability and accuracy.

[0097] In steps 202-204, the boundary layer mesh concentrates computing resources on the wall areas with the most intense physical gradients, ensuring the accuracy of the core heat transfer and flow resistance calculations. The dynamic mesh enables precise simulation of the functional movements of the SMA microactuator, and the static mesh with the lowest computational cost is applied to most non-moving areas of the model, greatly reducing the computational burden of subsequent simulations.

[0098] 205. In the mesh model, corresponding to the cooling fluid of the liquid cooling radiator, the heat dissipation substrate and the SMA microactuator, the fluid domain, the solid domain and the actuator domain are defined respectively;

[0099] In this embodiment, step 205 is similar to step 103 in the aforementioned embodiment and will not be described again here.

[0100] 206. Obtain phase change parameters of the SMA microactuator, where the material characteristic parameters include at least Young's modulus, phase change stress influence coefficient, and theoretical phase change temperature;

[0101] 207. Conduct differential scanning calorimetry experiments on samples made from the same batch of materials as the SMA microactuator to determine the precise phase transition temperature of the sample under actual thermal cycling. Based on the precise phase transition temperature, the theoretical phase transition temperature in the phase transition parameters is corrected, and the SMA constitutive model is established based on the corrected phase transition parameters.

[0102] In steps 206-207, to construct the SMA constitutive model, the first step is to obtain the phase transformation parameters of the SMA microactuator. These are physical constants that quantitatively describe the martensitic and reverse transformation behaviors of shape memory alloys under different temperature and stress conditions. These phase transformation parameters include, but are not limited to, Young's modulus, phase transformation stress influence coefficient, and theoretical phase transformation temperature. They also include Poisson's ratio, thermal expansion coefficient, latent heat of phase transformation, and the start and end temperatures of the martensitic and austenitic phase transformations. For specific information, the standard physical and mechanical properties of the corresponding SMA material can be found in the product datasheets provided by the material supplier.

[0103] Actual material properties can vary due to differences in production batches and processing. Phase transition temperatures, in particular, are highly sensitive to thermal cycling history. If a model built directly using theoretical parameters without experimental calibration is used, the predicted actuator operating temperature may differ significantly from actual conditions, resulting in distortion in the digital twin model. Therefore, to improve the accuracy of the SMA constitutive model, experimental testing of the actual material is necessary to obtain the most realistic phase transition temperature. This information can then be used to refine the model and ensure that the simulation results closely reflect the actual behavior of the physical prototype. Specifically, a sample made from the same batch of material as the SMA microactuator is placed in a differential scanning calorimeter (DSC). The instrument performs a complete heating and cooling cycle on the sample and outputs a DSC curve. This curve indicates the phase transition process as distinct endothermic or exothermic peaks. By analyzing these peaks, the actual phase transition onset and end temperatures can be precisely determined. Subsequently, the SMA constitutive model is established using this set of corrected phase change parameters, which can improve the accuracy of the SMA constitutive model and ensure that the simulation results of the digital twin model can reflect the real behavior to the greatest extent.

[0104] Specifically, the SMA constitutive model architecture is either the Brinson or Auricchio architecture. The core feature of the Brinson architecture is the decomposition of the martensite volume fraction ξ into two components: stress-induced martensite ξS and temperature-induced martensite ξT. This decomposition allows for a clearer distinction and description of the material's pseudoelastic and shape memory effects within a unified framework. Its phase transformation kinetic equations can use cosine functions to describe the onset, progression, and end of the phase transformation process, offering advantages in describing certain specific phase transformation behaviors. The Auricchio architecture, established within a rigorous thermodynamic framework, offers excellent generalizability and numerical stability, making it particularly suitable for finite element calculations under complex three-dimensional stress states. It generally does not explicitly distinguish between stress-induced and temperature-induced martensite, but instead describes the phase transformation state through the martensite volume fraction. Its phase transformation kinetic evolution equations often use exponential functions, which provide smoother transitions in numerical calculations and facilitate solver convergence. The choice between the two depends on the specific application scenario, the implementation of the simulation software used, and which model matches the existing material parameters better. For the simulation in this embodiment, both can effectively capture its core phase change driving behavior.

[0105] 208. Based on the SMA constitutive model, a bidirectional coupling relationship is established between the fluid domain, solid domain and actuator domain to form a digital twin model. The bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling and fluid-solid coupling;

[0106] In this embodiment, step 208 is similar to step 104 in the above embodiment and will not be described in detail here. The following describes the establishment methods of conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling respectively:

[0107] For conjugate heat transfer coupling, specifically, a two-way heat transfer calculation is performed between all contact surfaces of the fluid domain, solid domain, and actuator domain to obtain the temperature field and heat flux density distribution in order to establish conjugate heat transfer coupling. By establishing conjugate heat transfer coupling, the software automatically processes the heat exchange on these contact surfaces, ensuring that the wall temperature and heat flux density are continuous and conserved on both sides of the interface during the solution process. It can accurately calculate the temperature distribution on the complete path from the source, through the substrate, and finally to the SMA microactuator, and accurately predict the real-time temperature of the SMA microactuator. This real-time temperature is the only signal that activates the phase change of the SMA microactuator and generates action. Therefore, the conjugate heat transfer coupling directly determines whether the entire digital twin model can respond correctly at the right time.

[0108] For thermal-deformation coupling, the temperature field is applied as a thermal load to the actuator domain. The phase change and resulting phase change strain caused by temperature changes are calculated using the SMA constitutive model. The macroscopic deformation displacement of the SMA microactuator is solved to establish thermal-deformation coupling. In other words, the temperature field required by the user-specified structural mechanics module in the actuator domain is directly derived from the calculation results of the thermal module. When the solver is running, the temperature value at each grid node is automatically transferred as a thermal load to the structural mechanics calculation and SMA constitutive model calculation at that node. Based on this temperature input, the SMA constitutive model calculates the corresponding phase change strain, thereby driving structural deformation.

[0109] Fluid-structure interaction specifically updates the geometric boundaries of the fluid domain in real time based on macroscopic deformation and displacement. Simultaneously, the pressure and shear forces generated by the fluid domain flowing over the surface of the SMA microactuator are fed back as mechanical loads into the structural mechanics calculations of the actuator domain to establish fluid-structure interaction. Specifically, the mesh deformation of the fluid domain is driven by the boundary displacements calculated by the structural mechanics module, while the loads acting on the actuator surface are derived from the pressure and shear forces calculated by the fluid mechanics module. Fluid-structure interaction directly links the mechanical motion of the SMA microactuator to pressure drop and energy consumption. When the SMA microactuator's motion changes the flow path geometry, the flow field is immediately redistributed, altering the total pressure drop, pumping power, and heat transfer efficiency. These changes, in turn, affect the temperature field, creating a new thermal state input for the SMA microactuator itself and establishing a complete heat dissipation cycle.

[0110] 209. Apply a time-varying simulated heat load to the digital twin model to obtain the simulation performance parameters of the liquid cooling radiator under dynamic conditions;

[0111] 210. Iteratively optimize the target design parameters according to the simulation performance parameters and the preset optimization target to obtain the target design parameters, wherein the target design parameters include at least one of the geometric size, shape, anchor point position and material phase change activation temperature of the SMA microactuator;

[0112] In this embodiment, steps 209-210 are similar to steps 105-106 in the aforementioned embodiment and are not described again here.

[0113] In some specific embodiments, iterative optimization requires setting a preset optimization target. By setting an optimization target of minimizing the fluid pumping power of the liquid-cooled radiator under low thermal load conditions, the problem of high energy consumption of existing microchannel radiators under low loads can be solved. By reducing the pumping power at low loads, the overall energy efficiency of the system can be significantly improved and operating costs can be reduced. By setting an optimization target of minimizing the response time of the SMA microactuator from on to off or from off to on, it can be ensured that the radiator can respond quickly and promptly to changes in the chip's thermal load. A response time that is too long will cause the system to be unable to quickly switch to the appropriate heat dissipation mode when the thermal load suddenly changes, affecting its performance and reliability. The fast-responding SMA microactuator can more effectively achieve dynamic adjustment of the heat dissipation intensity.

[0114] In some specific embodiments, after obtaining the simulated performance parameters of the liquid-cooled radiator under dynamic working conditions, the simulated performance parameters can be further compared with the benchmark performance parameters measured through physical prototype experiments to obtain performance deviations, and an inverse correction algorithm is used to iteratively adjust one or more phase change parameters in the SMA constitutive model until the performance deviation meets the preset requirements.

[0115] Despite material-level DSC experimental calibration, there may still be deviations from the macroscopic performance of the physical prototype due to factors that cannot be fully reproduced in the model, such as microscopic assembly stress, local material inhomogeneity, and slight differences in the actual heat dissipation environment. Therefore, at least one physical prototype can be manufactured according to an optimized design scheme determined in step 106. A dynamic heat load identical to that in the simulation in step 105 is applied on the constructed experimental platform, and the baseline performance parameters of the physical prototype are recorded. The baseline performance parameter curve measured in the physical experiment is compared with the performance parameter curve obtained in the simulation. By calculating the difference between the two curves, a quantitative performance deviation is obtained. If the performance deviation exceeds the preset allowable range, the inverse correction algorithm is activated to adjust the material phase change parameters within the SMA constitutive model, and then the digital twin model is driven to perform a complete transient simulation again to obtain a new set of simulation performance parameters. The deviation between the new simulation results and the experimental results is then calculated again, and the next parameter adjustment is determined based on the deviation. This process is repeated until the performance deviation is reduced to within the preset requirements, eliminating the gap between the theoretical model and the actual situation.

[0116] 211. Configure the damage evolution module for the SMA constitutive model. The damage evolution module is used to dynamically modify the phase transformation parameters according to the number of thermodynamic cycles.

[0117] 212. During the simulation process, the SMA constitutive model is continuously updated according to the damage evolution module, and the performance indicators of the SMA microactuator are monitored to verify whether the attenuation of the performance indicators meets the design redundancy range after reaching the preset cycle life number.

[0118] Because shape memory alloys experience tens of thousands of thermal cycles, their material properties do not remain constant but rather slowly degrade, a phenomenon known as functional fatigue. For example, their phase transition temperature may drift, and the recoverable deformation may decrease. If this failure behavior is not considered, the optimized design based on the initial parameters may not meet performance requirements after a period of actual product use. Therefore, in steps 211-212, a damage evolution module can be configured to simulate the decline of material properties over service life in the digital twin model, thereby predicting the long-term reliability of the product.

[0119] Specifically, a mathematical model describing the functional fatigue of the SMA is selected, and the relationship between phase transformation parameters and the number of cycles, N, is established. The damage evolution module can be based on an empirical or semi-empirical phenomenological model, often using a logarithmic or power-law function. This damage evolution module dynamically modifies the core phase transformation parameters in the SMA constitutive model, including the phase transformation temperature, maximum residual strain, and the phase transformation stress influence coefficient. The damage evolution module can be used to simulate long-term use to observe the specific extent of performance degradation, quantitatively assessing the performance of the SMA microactuator at the end of its design life, and ultimately determining whether this degradation will lead to product failure. The goal is to verify the long-term reliability of the design and ensure that the product maintains its basic functionality throughout its lifecycle, even with performance degradation. Throughout the simulation process, key performance indicators are continuously recorded as a function of the number of cycles. When the simulation reaches the preset number of cycles, the performance indicator values ​​at this point are compared with the minimum acceptable performance threshold set at the beginning of the design to determine whether they are still within the design margin. The core objective is to verify the maximum displacement of the SMA microactuator under actuation, verifying whether it is still sufficient to completely close the bypass channel at the end of its lifespan. For example, if the design requires an SMA microactuator stroke of at least 0.5mm to ensure sealing, and the initial design stroke is 0.7mm, the design margin is 0.2mm. This step verifies that the stroke attenuation is less than 0.2mm after N cycles. The phase change response temperature also needs to be verified to verify that its value after drift is still within the acceptable operating temperature range.

[0120] In this embodiment, an SMA microactuator is integrated into the liquid-cooled heat sink microchannel by designing a bypass channel parallel to the main channel. The optimized SMA microactuator design enables self-sensing of thermal load and adaptive adjustment of heat dissipation intensity within the compact heat sink microchannel, giving the heat sink the ability to detect and respond to temperature. The success of the heat sink's adaptive behavior depends crucially on precisely matching parameters such as the SMA microactuator's geometry, anchor point location, and phase change activation temperature with the thermal characteristics of the object being cooled. Therefore, a complete digital twin model encompassing conjugate heat transfer, thermal deformation, and fluid-structure coupling is established, and iterative optimization is performed based on this model to ensure that the self-sensing and self-response mechanisms within the resulting physically manufactured heat sink are optimal solutions that have been quantitatively designed and verified.

[0121] See also Figure 3 , Figure 3 An embodiment of a liquid-cooled radiator microchannel construction device with integrated bypass control provided in this application includes:

[0122] A first modeling unit 301 is configured to establish a three-dimensional geometric model of the liquid-cooled radiator, the three-dimensional geometric model including a main channel for processing peak heat load conditions, a bypass channel for processing low heat load conditions, and an SMA microactuator disposed in the bypass channel;

[0123] The mesh unit 302 is used to perform meshing processing on the three-dimensional geometric model for finite element numerical simulation to obtain a mesh model;

[0124] A definition unit 303 is used to define the fluid domain, solid domain, and actuator domain corresponding to the cooling fluid, heat dissipation substrate, and SMA microactuator of the liquid-cooled radiator in the grid model, and configure an SMA constitutive model for the actuator domain to describe the relationship between temperature, stress, and phase change;

[0125] The second modeling unit 304 is used to establish a bidirectional coupling relationship between the fluid domain, the solid domain, and the actuator domain based on the SMA constitutive model to form a digital twin model. The bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling.

[0126] The simulation unit 305 is used to apply a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid cooling radiator under dynamic working conditions;

[0127] The optimization unit 306 is configured to iteratively optimize the target design parameters according to the simulation performance parameters to obtain the target design parameters, where the target design parameters include at least one of the geometric size, shape, anchor point position, and material phase change activation temperature of the SMA microactuator.

[0128] Optionally, the second modeling unit 304 is specifically configured to:

[0129] Perform bidirectional heat transfer calculations between all contact surfaces of the fluid domain, solid domain, and actuator domain to obtain the temperature field and heat flux density distribution to establish conjugate heat transfer coupling;

[0130] The temperature field is applied to the actuator domain as a thermal load. The phase change and resulting phase change strain caused by temperature change are calculated using the SMA constitutive model. The macroscopic deformation displacement of the SMA microactuator is obtained to establish thermal-mechanical deformation coupling.

[0131] The geometric boundaries of the fluid domain are updated in real time according to the macroscopic deformation displacement. At the same time, the pressure and shear force generated when the fluid domain flows through the surface of the SMA microactuator are fed back as mechanical loads to the structural mechanics calculation of the actuator domain to establish fluid-structure coupling.

[0132] Optionally, the grid unit 302 is specifically configured to:

[0133] Generate multi-layer refined boundary layer meshes at the interface between the fluid region and all solid walls;

[0134] The area far away from the SMA microactuator is meshed using static mesh;

[0135] The SMA microactuator area and the fluid area adjacent to the SMA microactuator are meshed using dynamic deformation mesh or overlapping mesh.

[0136] Optionally, the second modeling unit 304 is further configured to:

[0137] Obtaining the phase transition parameters of the SMA microactuator, wherein the material characteristic parameters include at least Young's modulus, phase transition stress influence coefficient, and theoretical phase transition temperature;

[0138] Differential scanning calorimetry experiments were conducted on samples made from the same batch of materials as the SMA microactuator to determine the precise phase transition temperature of the sample under actual thermal cycling. The theoretical phase transition temperature in the phase transition parameters was corrected based on the precise phase transition temperature, and the SMA constitutive model was established based on the corrected phase transition parameters.

[0139] The simulated performance parameters are compared with the benchmark performance parameters measured through physical prototype experiments to obtain the performance deviation. The inverse correction algorithm is used to iteratively adjust one or more phase change parameters in the SMA constitutive model until the performance deviation meets the preset requirements.

[0140] Optionally, the model architecture of the SMA constitutive model is a Brinson architecture or an Auricchio architecture.

[0141] Optionally, the apparatus further includes an evolving damage unit 307, configured to:

[0142] A damage evolution module is configured for the SMA constitutive model. The damage evolution module is used to dynamically modify the phase transformation parameters according to the number of thermodynamic cycles.

[0143] During the simulation process, the SMA constitutive model is continuously updated according to the damage evolution module, and the performance indicators of the SMA microactuator are monitored;

[0144] Verify whether the attenuation of performance indicators meets the design redundancy range after reaching the preset cycle life number.

[0145] Optionally, the optimization unit 306 is specifically configured to:

[0146] Iteratively optimize the target design parameters according to the simulation performance parameters and the preset optimization targets to obtain the target design parameters;

[0147] The preset optimization objectives include at least minimizing the fluid pumping power of the liquid cooling radiator under low heat load conditions and minimizing the response time of the SMA microactuator from open to closed or from closed to open.

[0148] In the device of this embodiment, the functions of each unit are the same as those described above. Figure 1 or Figure 2 The steps in the method embodiment shown correspond to each other and will not be repeated here.

[0149] This application also provides a liquid cooling radiator microchannel construction system with integrated bypass control, please refer to Figure 4 , Figure 4 An embodiment of a liquid-cooled radiator microchannel construction system with integrated bypass control provided by the present application includes:

[0150] Processor 401, memory 402, input and output unit 403, bus 404;

[0151] The processor 401 is connected to the memory 402, the input and output unit 403 and the bus 404;

[0152] The memory 402 stores a program, and the processor 401 calls the program to execute any of the above methods for constructing a liquid-cooled radiator microchannel with integrated bypass control.

[0153] The present application also relates to a computer-readable storage medium having a program stored thereon. When the program is run on a computer, the computer is caused to execute any of the above methods for constructing a liquid-cooled radiator microchannel with integrated bypass control.

[0154] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0155] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0156] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0157] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0158] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk, and other media that can store program code.

Claims

1. A method for constructing a liquid cooling radiator microchannel with integrated bypass control, characterized in that: The liquid cooling radiator microchannel construction method comprises: Establishing a three-dimensional geometric model of a liquid-cooled radiator, the three-dimensional geometric model includes a main channel for handling peak thermal load conditions, a bypass channel for handling low thermal load conditions, and an SMA microactuator disposed in the bypass channel. The SMA microactuator is made of a shape memory alloy material and is a solid component with a specific geometric shape. The SMA microactuator is configured to respond to temperature changes. When the coolant temperature is below a phase change activation point, the SMA microactuator remains stable in a preset low-temperature state, allowing the bypass channel to fully open. When the coolant temperature reaches the phase change activation point, the SMA microactuator undergoes a phase change and drives itself back to the memorized high-temperature state, closing or completely sealing the inlet of the bypass channel and forcing all the coolant into the main channel. Performing mesh generation processing for finite element numerical simulation on the three-dimensional geometric model to obtain a mesh model; In the grid model, corresponding to the cooling fluid of the liquid-cooled radiator, the heat dissipation substrate, and the SMA microactuator, a fluid domain, a solid domain, and an actuator domain are defined respectively, and an SMA constitutive model for describing the relationship between temperature, stress, and phase change is configured for the actuator domain; establishing a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model to form a digital twin model, wherein the bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling; Applying a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions; The target design parameters are iteratively optimized according to the simulation performance parameters to obtain the target design parameters, wherein the target design parameters include at least one of the geometric size, shape, anchor point position and material phase change activation temperature of the SMA microactuator.

2. The method for constructing a liquid cooling radiator microchannel according to claim 1, characterized in that: The establishing of a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model includes: performing a bidirectional heat transfer calculation among all contact surfaces of the fluid domain, the solid domain, and the actuator domain to obtain a temperature field and a heat flux density distribution to establish a conjugate heat transfer coupling; Applying the temperature field as a thermal load to the actuator domain, calculating the phase change and the resulting phase change strain caused by the temperature change through the SMA constitutive model, and solving to obtain the macroscopic deformation displacement of the SMA microactuator to establish thermal-mechanical deformation coupling; The geometric boundaries of the fluid domain are updated in real time according to the macroscopic deformation displacement. At the same time, the pressure and shear force generated when the fluid domain flows through the surface of the SMA microactuator are fed back as mechanical loads to the structural mechanics calculation of the actuator domain to establish fluid-solid coupling.

3. The method for constructing a liquid cooling radiator microchannel according to claim 1, characterized in that: The meshing process for finite element numerical simulation is performed on the three-dimensional geometric model to obtain a mesh model, including: Generate multi-layer refined boundary layer meshes at the interface between the fluid region and all solid walls; Meshing the area away from the SMA microactuator using a static grid; The region of the SMA microactuator and the fluid region adjacent to the SMA microactuator are meshed using a dynamic deformation grid or an overlapping grid.

4. The method for constructing a liquid cooling radiator microchannel according to claim 1, wherein: The SMA constitutive model configured for the actuator domain to describe the relationship between temperature, stress and phase change includes: Obtaining phase change parameters of the SMA microactuator, wherein the phase change parameters include at least Young's modulus, phase change stress influence coefficient, and theoretical phase change temperature; Performing a differential scanning calorimetry experiment on a sample made from the same batch of materials as the SMA microactuator to determine the precise phase transition temperature of the sample under actual thermal cycling, and correcting the theoretical phase transition temperature in the phase transition parameters based on the precise phase transition temperature, and establishing an SMA constitutive model based on the corrected phase transition parameters; After applying a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions, the method further includes: The simulated performance parameters are compared with the benchmark performance parameters measured through physical prototype experiments to obtain performance deviations, and an inverse correction algorithm is used to iteratively adjust one or more phase change parameters in the SMA constitutive model until the performance deviation meets the preset requirements.

5. The method for constructing a liquid cooling radiator microchannel according to claim 4, characterized in that: The model architecture of the SMA constitutive model is a Brinson architecture or an Auricchio architecture.

6. The method for constructing a liquid cooling radiator microchannel according to claim 4, characterized in that: The liquid cooling radiator microchannel construction method further includes: Configuring a damage evolution module for the SMA constitutive model, wherein the damage evolution module is used to dynamically modify the phase change parameters according to the number of thermodynamic cycles; During the simulation process, the SMA constitutive model is continuously updated according to the damage evolution module, and the performance indicators of the SMA microactuator are monitored; Verify whether the attenuation of the performance indicator meets the design redundancy range after reaching the preset cycle life number.

7. The method for constructing a liquid cooling radiator microchannel according to any one of claims 1 to 6, characterized in that: The iteratively optimizing the target design parameters according to the simulation performance parameters to obtain the target design parameters includes: Iteratively optimizing the target design parameters according to the simulation performance parameters and the preset optimization targets to obtain the target design parameters; The preset optimization objectives include at least minimizing the fluid pumping power of the liquid-cooled radiator under low heat load conditions and minimizing the response time of the SMA microactuator from open to closed or from closed to open.

8. A liquid cooling radiator microchannel construction device with integrated bypass control, characterized in that: The device comprises: a first modeling unit, configured to establish a three-dimensional geometric model of a liquid-cooled radiator, the three-dimensional geometric model comprising a main channel for handling peak heat load conditions, a bypass channel for handling low heat load conditions, and an SMA microactuator disposed within the bypass channel, the SMA microactuator being made of a shape memory alloy material and being a solid component having a specific geometric shape, the SMA microactuator being configured to respond to temperature changes; when the coolant temperature is lower than a phase change activation point, the SMA microactuator being stabilized in a preset low-temperature state, such that the bypass channel is fully open; and when the coolant temperature reaches the phase change activation point, the SMA microactuator undergoes a phase change and drives itself back to a memorized high-temperature state, such that the inlet of the bypass channel is closed or completely sealed, and all the coolant is forced into the main channel; A grid unit is used to perform a grid division process on the three-dimensional geometric model for finite element numerical simulation to obtain a grid model; defining a unit for defining a fluid domain, a solid domain, and an actuator domain in the grid model corresponding to the cooling fluid, the heat dissipation substrate, and the SMA microactuator of the liquid-cooled radiator, respectively, and configuring an SMA constitutive model for describing the relationship between temperature, stress, and phase change for the actuator domain; a second modeling unit, configured to establish a bidirectional coupling relationship among the fluid domain, the solid domain, and the actuator domain according to the SMA constitutive model to form a digital twin model, wherein the bidirectional coupling relationship includes conjugate heat transfer coupling, thermal deformation coupling, and fluid-solid coupling; A simulation unit, configured to apply a simulated heat load that varies with time to the digital twin model to obtain simulation performance parameters of the liquid-cooled radiator under dynamic working conditions; An optimization unit is configured to iteratively optimize target design parameters according to the simulation performance parameters to obtain target design parameters, wherein the target design parameters include at least one of the geometric size, shape, anchor point position, and material phase change activation temperature of the SMA microactuator.

9. A liquid cooling radiator microchannel construction system with integrated bypass control, characterized in that: The system comprises: processor, memory, input and output units, and buses; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, and when the program is executed on a computer, the method according to any one of claims 1 to 7 is executed.

Images