High-flux heat dissipation expandable topological flow channel and design method thereof

Abstract

The invention discloses a high-flux heat dissipation expandable topological flow channel and a design method thereof. The flow channel comprises a large inlet structure, a large diffusion structure and a large outlet structure. The large inlet structure is a large-area liquid inlet; the topological micro-channel with the large diffusion structure is in a three-section type and comprises a front section, a middle section and a rear section, fins of the front section introduce an inflowing cooling working medium to the two sides from a secondary channel of the front section and are matched with first middle fins of a main channel of the front section to draw close the cooling working medium to the middle, and the main channel of the front section is gradually narrowed; the second middle fins of the middle section divide the cooling working medium in the main channel of the front section into first edge fluid, and the first edge fluid gathers towards the middle under the influence of the fins of the middle section and is divided into second edge fluid by the shunting fins of the middle section; the second edge fluid is gathered at the rear section, and the heat exchange area is increased by utilizing the third middle fins at the rear section; and the large outlet structure is an outlet for discharging liquid in a large area.

Description

Technical Field

[0001] The present invention belongs to the technical field of chip thermal management, and in particular relates to an expandable topological flow channel for high-throughput heat dissipation and a design method thereof. Background Art

[0002] With the development of integrated circuit manufacturing and chip packaging technology and people's demand for higher computing power chips, the number of transistors per unit area of ​​the chip has greatly increased, resulting in a rapid increase in chip power density, and the local heat flux density can reach 1000W / cm 2 The temperature of the chip continues to rise during operation. If the heat cannot be dissipated quickly, the chip power consumption will increase, the chip multi-core operating frequency will be limited, the chip aging will be accelerated, and the chip will fail or even burn out. Studies have shown that more than 95% of the space in an electronic chip is used for heat dissipation, and the temperature of an electronic chip that works stably and continuously cannot exceed 85°C. With the further improvement of the degree of chip integration, higher requirements are placed on heat dissipation. The further development of chip technology requires more advanced thermal management technology as support.

[0003] At present, the research on chip microchannel cooling methods is concentrated on manifold microchannels, jet microchannels and patch-type external microchannels. Although these cooling methods have shown relatively strong cooling performance, they still have many shortcomings, such as poor stability of working fluid flow, uneven heat transfer leading to the generation of local hot spots, etc., which restricts the further improvement of chip cooling level.

[0004] Traditional chip microchannel cooling adopts a parallel structure of multiple horizontal microchannels, which has low processing precision requirements and is not easy to clog, but it also has disadvantages such as unstable flow, poor thermal uniformity, and high pump work. The flow rate between different flow channels is quite different, and the fluid flow rate in the central flow channel is much higher than that in the edge flow channel. The long process leads to a large pressure drop, and the mismatch between the pressure gradient and the temperature gradient leads to low heat transfer efficiency. As the chip computing power increases, the heat flux density continues to increase, and the heat dissipation capacity of traditional flat microchannels can no longer match it, which has limited the further development of chip microchannel cooling. Topological microchannel structure optimization has emerged as an emerging microchannel design method.

[0005] Compared with traditional straight microchannels, topological microchannels have greatly improved flow stability and thermal uniformity, and have gradually become an important means to solve the heat dissipation of high-power density chips. The topological microchannel structure optimizes the microchannel structure based on fluid mechanics, structural mechanics, and fractal geometry, which can reduce the flow resistance and heat transfer resistance of the working fluid, improve flow stability and thermal uniformity, and significantly improve the heat dissipation performance of the microchannel, especially for large-size and high-heat flux density chips. With the continuous development of lithography technology and microchannel fine processing technology, it has become possible to directly construct topological microchannels on the chip substrate, which can further improve the heat dissipation capacity of chip microchannels.

[0006] The traditional topological channel design for large-size heat dissipation is often based on the form of "small inlet-large diffusion-small outlet", while the actual chip heat dissipation process is mainly based on "large inlet-large diffusion-large outlet". The traditional large-area topological channel design method deviates from the actual heat dissipation conditions; in addition, the traditional topological channel design has large dead areas at the corners, which easily accumulates heat and causes hot spots. It also has the disadvantages of long calculation time and the pursuit of fine bionic structures is prone to blockage, resulting in low practical application value. Summary of the invention

[0007] The present invention aims to solve the deficiencies of the prior art and provides the following solutions:

[0008] A scalable topological flow channel for high-flux heat dissipation, comprising: a large inlet structure, a large diffusion structure and a large outlet structure;

[0009] The large inlet structure is an inlet for liquid inlet over a large area;

[0010] The topological microfluidic channel of the large diffusion structure is a three-section type, including: a front section, a middle section and a rear section;

[0011] The front section includes a first axial fin and five first middle fins. The front section starts from the first axial fin. The first axial fin is arranged on the first symmetry axis. The five first middle fins are arranged in sequence and close to the second symmetry axis. Gaps are provided between the first axial fin and the five first middle fins. The fins of the front section introduce the inflowing cooling medium from the front section secondary channel to both sides, and the first middle fins cooperate with the front section main channel to bring the cooling medium close to the middle. The front section main channel gradually narrows.

[0012] The middle section includes a second axial fin, a third axial fin, a fourth axial fin and three second middle fins. The middle section starts from the third axial fin. The third axial fin is arranged on the second symmetry axis. The three second middle fins are arranged in sequence and approach the first symmetry axis to reach the second axial fin on the first symmetry axis. At the same time, the fourth axial fin is arranged at the end position of the middle section. Gaps are provided between the second axial fin, the third axial fin, the fourth axial fin and the three second middle fins. The second middle fin of the middle section divides the cooling medium in the main channel of the front section into a first edge fluid. The first edge fluid gathers toward the middle under the influence of the fins of the middle section and is divided into a second edge fluid by the diverter fins of the middle section.

[0013] The rear section includes three third middle fins, which are arranged in sequence from the second axial fins and close to the second symmetry axis, and gaps are provided between the three third middle fins; the second edge fluid is collected in the rear section, and the heat exchange area is increased by using the third middle fins of the rear section;

[0014] The large outlet structure is an outlet for discharging liquid over a large area.

[0015] The present invention also provides a design method for a high-flux heat dissipation expandable topology flow channel, the design method is used to design the above expandable topology flow channel, comprising the following steps:

[0016] According to the actual working environment of the chip heat sink, a dual-plane pseudo three-dimensional topology optimization model is established, and the physical parameters, heat source distribution and cooling medium parameters of the chip heat sink are obtained;

[0017] Determine the chip operating temperature, cooling fluid boundary conditions and heat source description form, determine the basic method of topology optimization problem based on the physical parameters of the chip heat sink, the heat source distribution and the cooling fluid parameters, and construct a topology optimization mathematical model by setting the topology optimization objective function;

[0018] Performing finite element meshing on the three-dimensional topology optimization model and the topology optimization mathematical model, initializing model-related physical fields, calculating to obtain a topology optimization model, and performing parameter iteration on the topology optimization model to obtain a final topology optimization model;

[0019] Based on the final topology optimization model, a spline function is used to fit a topology optimization two-dimensional model, and the topology optimization two-dimensional model is symmetrically expanded to obtain the expandable topology flow channel.

[0020] Preferably, the physical parameters include: radiator width W = 0.5 ~ 3mm, radiator length L2 = 5 ~ 20mm, radiator thickness z = 0.1 ~ 1mm, radiator inlet length L in =0.5~2mm, radiator outlet length L out =0.5~2mm, radiator material, radiator material thermal conductivity k s , Radiator material constant pressure specific heat capacity C p,s and heat sink material density ρ s ;

[0021] The heat source distribution includes: the heat source position is the α1 plane, and the application direction is vertically conducted from the α1 plane to the α2 plane;

[0022] The cooling medium parameters include: cooling medium composition, cooling medium heat conduction coefficient k f, constant pressure specific heat capacity of cooling medium C p,f and cooling medium density ρ f .

[0023] Preferably, the cooling medium boundary conditions include: inlet pressure p in =10~200Pa, outlet pressure p out =0Pa and inlet temperature T in =293.15~313.15K;

[0024] The heat source description form is:

[0025] F=h e (1-γ)(T s -T)

[0026] Where F represents the total heat transfer in the design domain, h e represents the heat transfer coefficient of the fin in the α1 plane, γ represents the topology optimization design variable, T s represents the chip operating temperature, and T represents the temperature in the design domain.

[0027] Preferably, the method for constructing the topology optimization mathematical model includes:

[0028] Based on the physical parameters of the chip heat sink, the distribution of the heat source and the parameters of the cooling medium, the variable density method is used to add laminar flow and fluid heat transfer physical fields to the pseudo three-dimensional topology optimization model, and the properties of the virtual solid-liquid mixed state in the design domain are defined by the material interpolation equation model:

[0029]

[0030] Among them, α(γ) represents the resistance coefficient, α f represents the reverse permeability in the fluid domain, α s represents the reverse permeability in the solid domain, ρ(γ) represents the virtual solid-liquid state density in the domain, C p (γ) represents the heat capacity of the virtual solid-liquid state in the domain, k(γ) represents the density thermal conductivity of the virtual solid-liquid state in the domain, q, q ρ ,q Cp and q k represents the optimization parameter, μ in represents the inlet fluid viscosity, Da represents the Darcy number;

[0031] Setting the total heat transfer F in the design domain as the topology optimization objective function;

[0032] Set the constrained solid phase fraction and build the topology optimization mathematical model:

[0033]

[0034]

[0035] Where Ψ represents the topology optimization design domain, V Ψ Design domain volume for topology optimization, V s represents the set liquid phase fraction, p represents the pressure, u represents the velocity of the fluid in the x direction, v represents the velocity of the fluid in the y direction, and μ represents the dynamic viscosity of the fluid.

[0036] Preferably, the method for obtaining the final topology optimization model includes:

[0037] Finite element meshing is performed on the three-dimensional topology optimization model and the topology optimization mathematical model, and the resulting mesh independence and checkerboard problems are solved using a Helmholtz filter:

[0038]

[0039] in, represents the auxiliary function after the topology optimization design variable γ is filtered by the Helmholtz filter, represents the Hamiltonian operator, r m Indicates the filtering parameters. Indicates that The function after hyperbolic tangent projection, γ β represents the optimization parameters;

[0040] Using the optimization iterative algorithm, by optimizing the parameters q, γ β and Da perform continuous processing on the filter function, set the maximum number of iteration steps and the iteration termination condition of each step, optimize the topology optimization objective function, and initialize the density field, temperature field and velocity field of the model to obtain the topology optimization model;

[0041] The topology optimization model is used as the initial value, and the optimization parameters are set to solve the next topology optimization model, until the final convergence requirement is reached to complete the entire continuous optimization process, and the final topology optimization model is output.

[0042] Preferably, the method for obtaining the expandable topological flow channel includes:

[0043] The final topology optimization model is fitted using a B-spline function to obtain a fitted topology optimization two-dimensional model;

[0044] The topology optimization two-dimensional model uses the bilateral symmetry setting in the boundary conditions, takes the bilateral symmetric boundaries as the symmetry axis, repeats the symmetric expansion multiple times, deletes the overlapping original symmetric condition boundaries, converts the remaining symmetric boundary conditions into internal boundary conditions, and obtains the expandable topology flow channel.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention is based on a variable density topology optimization method. By modifying the optimization model to match the actual heat dissipation process and reasonably designing the optimization boundary conditions to make it scalable, the poor thermal conductivity of the corners of the traditional topology optimization model is avoided together. The topological channel designed by this method significantly improves the heat dissipation capacity, solves the problems of mismatch between traditional topology optimization and the actual heat dissipation process, local dead zones prone to local hot spots and poor scalability, and the scalability is suitable for the actual preparation of heat sinks for large-size chips. Compared with straight channels, the heat exchange capacity is significantly improved, and laser etching is actually used for processing and preparation. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0048] Figure 1 A schematic diagram of a structure of an expandable topological flow channel in an embodiment of the present invention;

[0049] Figure 2 A schematic diagram of a large diffusion structure in an expandable topological flow channel structure according to an embodiment of the present invention;

[0050] Figure 3 A schematic diagram of a method flow chart of an embodiment of the present invention;

[0051] Figure 4 It is a schematic diagram of the structure of a dual-plane pseudo three-dimensional topology optimization model according to an embodiment of the present invention;

[0052] Figure 5 This is a schematic diagram of the structure of a large-size chip heat sink model according to an embodiment of the present invention;

[0053] Figure 6 This is a typical silicon embedded expandable topological channel obtained by laser etching in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Embodiment 1

[0057] In this embodiment, if Figure 1 , Figure 2 As shown, a scalable topological flow channel for high-throughput heat dissipation includes: a large inlet structure, a large diffusion structure and a large outlet structure.

[0058] The large inlet structure is an inlet with a large liquid inlet area.

[0059] The topological microchannel of the large diffusion structure is a three-section type, including: a front section, a middle section and a rear section; the front section includes a first axial fin and five first middle fins, the front section starts from the first axial fin, the first axial fin is arranged on the first symmetry axis, the five first middle fins are arranged in sequence and approach the second symmetry axis, and gaps are provided between the first axial fins and the five first middle fins; the fins of the front section introduce the incoming cooling medium from the front section secondary channel to both sides, and cooperate with the first middle fins of the front section main channel to bring the cooling medium closer to the middle, and the front section main channel gradually narrows; the middle section includes a second axial fin, a third axial fin, a fourth axial fin and three second middle fins, the middle section starts from the third axial fin, the third axial fin is arranged on the second symmetry axis, and the three second middle fins are arranged on the second symmetry axis. The fins on the first section are arranged in sequence and approach the first symmetry axis to reach the second axial fin on the first symmetry axis. At the same time, the fourth axial fin is arranged at the end position of the middle section, and gaps are provided between the second axial fin, the third axial fin, the fourth axial fin and the three second middle fins; the second middle fin in the middle section divides the cooling medium in the main channel of the front section into a first edge fluid, and the first edge fluid gathers toward the middle under the influence of the fins in the middle section, and is divided into a second edge fluid by the diverter fins in the middle section; the rear section includes three third middle fins, and the three third middle fins are arranged in sequence starting from the second axial fin and approach the second symmetry axis, and gaps are provided between the three third middle fins; the second edge fluid is gathered in the rear section, and the third middle fin in the rear section is used to increase the heat exchange area.

[0060] In this embodiment, the spacing between the fins is 0.1-0.2mm, the spacing between the fins on the shaft and the middle fins is larger, the three sections of fins are evenly distributed along the process horizontally, the fin lengths are all about 0.5mm, and the area of ​​the middle fins will be significantly increased when they are close to the fins on the shaft; the longitudinal direction is stepped close to the symmetry axes on both sides, and the spacing between the fins is similar; the front end has a larger number of fins for liquid distribution, which increases the number of multi-stage channels, slows down the liquid distribution flow rate, and makes the liquid distribution more uniform; the middle and rear fins play a greater role in disturbing the fluid.

[0061] The channels formed between the axial fins and the middle fins are main channels, that is, the channels with larger width along the flow direction are the main channels, and there are 2 main channels in total; the channels formed between the middle fins are secondary channels, that is, the channels with narrower widths perpendicular to the flow direction are the secondary channels, and there are 9 main channels and secondary channels in total, which are basically in a vertical relationship; the 5 first middle fins in the front section form 4 secondary channels, which are evenly distributed laterally and tend to narrow along the flow process; the 3 second middle fins in the middle section form 2 secondary channels with similar widths; the 3 third middle fins in the rear section form channels with similar widths; after multi-fin diversion, it finally presents a shape of multiple secondary channels evenly distributed between the two main channels, and the main channels and secondary channels are intertwined to realize the overall flow process of liquid distribution-multiple liquid separation-outflow.

[0062] The large outlet structure is an outlet with a large liquid discharge area.

[0063] In this embodiment, the overall model setting of "large inlet-large diffusion-large outlet" conforms to the current actual chip heat dissipation method. The "large inlet" is a large inlet structure, and the inlet area of ​​the inlet area is large, so that the cooling medium has a larger flow rate at the same flow rate, and the liquid distribution is more uniform, which can significantly improve the heat dissipation capacity and temperature uniformity of the cooling medium;

[0064] "Large diffusion" and large diffusion structure refer to the large area of ​​the cooling medium flowing and transferring heat in the design domain, and the fluid can fully flow and transfer heat in the domain, such as Figure 2 As shown in the figure, the topological microfluidic channel after symmetrical expansion along the second symmetry axis is formed in the domain and is divided into three sections, including the front section, the middle section and the back section. The front section fins introduce the central fluid from the secondary channel to both sides, and cooperate with the first middle fin of the main channel to move closer to the middle, so that the secondary channel can achieve more uniform large-area liquid distribution; the middle section further divides the central liquid into the first edge liquid through the middle second middle fin, and at the same time, the edge side liquid tends to gather in the middle, and the mutual convection forms a local vortex, which strengthens the disturbance between the fluids, and the two parts of the liquid are fully mixed. The larger main channel accelerates the flow rate of the fluid, and at the same time, the second middle fin of the middle section introduces part of the central fluid into the second edge fluid, and the second edge fluid is collected in the back section. At the same time, the fins increase the convection heat exchange area to avoid the temperature increase of the edge fluid in the back section, forming a basic process of uniform liquid distribution in the front section, strengthening the mixing and disturbance of the central fluid and the edge fluid in the middle section, and rapid outflow and further strengthening the disturbance in the back section, which perfectly meets the heat dissipation problem of large-size chips targeted by the proposal.

[0065] "Big outlet" refers to a large outlet structure with a large outlet area, which speeds up the flow process of the fluid working medium, while reducing the pressure drop in the outlet part, avoiding the problem of local backflow before the outlet of the "small outlet".

[0066] Embodiment 2

[0067] In this embodiment, if Figure 3As shown, a design method for a scalable topological flow channel with high-throughput heat dissipation includes the following steps:

[0068] S1. According to the actual working environment of the chip heat sink, a dual-plane pseudo three-dimensional topology optimization model is established, and the physical parameters, heat source distribution and cooling medium parameters of the chip heat sink are obtained.

[0069] In this embodiment, a heat sink physical model is first constructed and simplified into a dual-plane pseudo three-dimensional topology optimization model, such as Figure 4 As shown, it includes an α1 plane and an α2 plane, wherein the α1 plane is regarded as the bottom plane of the heat sink, and the α2 plane is regarded as the middle cross-section of the chip heat sink.

[0070] Physical parameters include: radiator width W = 0.5 ~ 3mm, radiator length L2 = 5 ~ 20mm, radiator thickness z = 0.1 ~ 1mm, radiator inlet length L in =0.5~2mm, radiator outlet length L out =0.5~2mm, radiator material, radiator material thermal conductivity k s , Radiator material constant pressure specific heat capacity C p,s and heat sink material density ρ s In this embodiment, the heat sink material can be copper or silicon; the heat source distribution includes: the heat source position is the α1 plane, and the application direction is vertically conducted from the α1 plane to the α2 plane; the cooling medium parameters include: cooling medium composition, cooling medium thermal conductivity coefficient k f , constant pressure specific heat capacity of cooling medium C p,f and cooling medium density ρ f In this embodiment, the cooling medium is water, acetone, HFE-7100, and ethanol.

[0071] S2. Determine the chip operating temperature, cooling fluid boundary conditions and heat source description form. Based on the physical parameters of the chip heat sink, heat source distribution and cooling fluid parameters, determine the basic method of the topology optimization problem, and construct a topology optimization mathematical model by setting the topology optimization objective function.

[0072] The boundary conditions of the cooling medium include: inlet pressure p in =10~200Pa, outlet pressure p out =0Pa and inlet temperature T in =293.15~313.15K;

[0073] The heat source is described in the form of α1 plane fin heat transfer coefficient h e Multiply it by the solid fraction in the design domain and then by the chip operating temperature T s And the product of the design domain temperature T difference:

[0074] F=h e (1-γ)(T s -T)

[0075] 0≤γ≤1

[0076] Where F represents the total heat transfer in the design domain, h e represents the heat transfer coefficient of the fin in the α1 plane, h e =10 5 ~10 8 W / (m 3 ·K), γ represents the topology optimization design variable, T s Indicates the chip operating temperature, T s =358.15K, T represents the temperature in the design domain, 0 represents the heat dissipation material (solid domain), and 1 represents the coolant (fluid domain).

[0077] In this embodiment, the method for constructing a topology optimization mathematical model includes:

[0078] Based on the physical parameters of the chip heat sink, the distribution of heat sources and the parameters of the cooling medium, the variable density method is used to add laminar flow and fluid heat transfer physical fields to the pseudo three-dimensional topology optimization model. The fluid flow is defined as incompressible flow. The heat transfer process only considers heat conduction and convection heat transfer, and does not consider thermal radiation and fluid viscous dissipation. The non-isothermal flow multi-physics field coupling interface is set, and the properties ρ(γ), C in the virtual solid-liquid mixed state in the design domain are defined through the material interpolation equation model. p (γ) and k(γ):

[0079]

[0080] Among them, α(γ) represents the resistance coefficient, α f represents the reverse permeability in the fluid domain, α f =0,α s represents the reverse permeability in the solid domain, ρ(γ) represents the virtual solid-liquid state density in the domain, C p (γ) represents the heat capacity of the virtual solid-liquid state in the domain, k(γ) represents the density thermal conductivity of the virtual solid-liquid state in the domain, q, q ρ ,q Cp and q k represents the optimization parameter, set to 0.01~0.05, μ in Indicates the inlet fluid viscosity, Da indicates the Darcy number, Da = 10 -6 ~10 -3 ;

[0081] The total heat transfer F in the design domain is set as the topology optimization objective function;

[0082] Set the constrained solid fraction and build a topology optimization mathematical model:

[0083]

[0084]

[0085] Where Ψ represents the topology optimization design domain, V Ψ Design domain volume for topology optimization, V s Indicates the set liquid fraction, V s =0.1~0.5, p represents pressure, u represents the velocity of the fluid in the x direction, v represents the velocity of the fluid in the y direction, and μ represents the dynamic viscosity of the fluid.

[0086] S3. Perform finite element meshing on the three-dimensional topology optimization model and the topology optimization mathematical model, initialize the model-related physical fields, calculate the topology optimization model, and iterate the parameters of the topology optimization model to obtain the final topology optimization model.

[0087] Methods to obtain the final topology optimization model include:

[0088] Finite element meshing is performed on the three-dimensional topology optimization model and the topology optimization mathematical model, and the Helmholtz filter is used to solve the mesh independence and checkerboard problems:

[0089]

[0090] in, represents the auxiliary function after the topology optimization design variable γ is filtered by the Helmholtz filter, represents the Hamiltonian operator, r m Indicates the filtering parameters. Indicates that The function after hyperbolic tangent projection, γ β represents the optimization parameters;

[0091] Using the optimization iterative algorithm, by optimizing the parameters q, γ β The filter function is processed continuously with Da, the maximum number of iteration steps and the iteration termination condition of each step are set, the topology optimization objective function is optimized, and the density field, temperature field and velocity field of the model are initialized to obtain the topology optimization model; in this embodiment, the moving progressive algorithm MMA or SNOPT is selected as the optimization iterative algorithm, and the parameters q, γ are used to adjust the optimization parameters. β The filter function is processed continuously with Da. Different optimization parameter values ​​are used for each step of the optimization process. The maximum number of iterations and the termination condition of each step are set. When |F i+1 -F i |≤n, where F iis the objective function value obtained in the i-th iteration, i is the number of iteration steps, the maximum number of iteration steps is set to 100, n is the upper limit of iteration convergence, and the parameter value settings are shown in Table 1:

[0092] Table 1

[0093]

[0094] Initialize the density field, temperature field and velocity field of the model, and obtain the topology optimization model after the first step of calculation. Use the topology optimization model as the initial value, set the optimization parameters to solve and obtain the next topology optimization model, until the final convergence requirement is reached to complete the entire continuous optimization process and output the final topology optimization model.

[0095] S4. Based on the final topology optimization model, the spline function is used for fitting to obtain the topology optimization two-dimensional model, and the topology optimization two-dimensional model is symmetrically expanded to obtain an expandable topology flow channel.

[0096] In this embodiment, the method for obtaining the expandable topological flow channel includes:

[0097] The final topology optimization model is imported into CAD, and the B-spline function is used to fit it to obtain the fitted topology optimization two-dimensional model; the topology optimization two-dimensional model uses the bilateral symmetry setting in the boundary conditions, takes the bilateral symmetric boundaries as the symmetry axis, repeats the symmetric expansion many times, deletes the overlapping original symmetric condition boundaries, converts the remaining symmetric boundary conditions into internal boundary conditions, obtains the expandable topology flow channel, and then symmetrically expands the expandable topology flow channel to obtain a large-size chip heat sink model, such as Figure 5 shown.

[0098] Then, the fitted expandable topology flow channel is imported into the CFD software, the fluid working medium and solid material properties are set, and relevant boundary conditions are added; according to the boundary conditions, the heat source characteristics are set, including the heat source action form, heat source location, and heat source size; the fluid working medium characteristics are set, including the inlet temperature T in , inlet pressure p in , outlet pressure p out , build a fluid-solid coupling heat transfer simulation model; mesh the model, set the convergence conditions and the maximum number of iterations, and select the Simple solution method for solution. Then change the fluid working medium inlet pressure p in and liquid fraction V s , and repeat the above steps to obtain various topological channels and the corresponding temperature and velocity fields, and calculate the relevant heat transfer data:

[0099]

[0100] T max =max(T1,T2,...,Tn )

[0101] Where R represents thermal resistance, T w represents the cross-sectional temperature, T ave represents the average temperature, A represents the mesh surface of the fluid-solid interface, T n Represents the nth grid temperature data, A n Represents the area of ​​the nth grid, and N represents the number of grids; based on the calculation results, it is compared with the straight channel, as shown in Table 2:

[0102] Table 2

[0103]

[0104] It can be seen from the table that the average temperature, maximum temperature and equivalent thermal resistance of the topological channel are significantly lower than those of the straight channel. The average temperature of the topological channel is 27.54K lower than that of the straight channel, the maximum temperature is 13.41K lower than that of the straight channel, and the thermal resistance is 0.737K·cm lower than that of the straight channel. 2 / W, the thermal resistance of the straight channel is 235.73% of that of the topological channel. It can be seen that the topologically optimized three-dimensional channel structure obtained by the method of the present invention is significantly superior to the traditional straight channel design in terms of heat dissipation capacity and temperature uniformity.

[0105] Embodiment 3

[0106] In this embodiment, a method for preparing the expandable topological flow channel is also provided:

[0107] Import the expandable topological flow channel structure into the CAD software and calculate the required channel area according to the required channel area. 2 Draw, reserve the anode bonding area, liquid inlet and liquid distribution area, fill the area to be etched, and obtain the two-dimensional CAD drawing to be etched.

[0108] Determine the relevant parameters such as etching laser intensity, etching method, etching time, etc. The processing equipment is a UV laser marking machine, using nanosecond UV etching technology, vertical bidirectional etching, laser intensity of 2~5W, pulse width of 1ns, pulse frequency of 40KHz, speed of 1000mm / a, and etching time of 1.5h~2.5h.

[0109] Place a two-inch silicon wafer with a thickness of 550μm in the laser etching processing area, and etch it according to the relevant setting parameters to obtain a typical topological channel structure with a depth of 300μm, such as Figure 6 As shown, it will be applied to the preparation of actual chip heat sinks later.

[0110] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A scalable topology channel for high-throughput heat dissipation, It is characterized in that include: Large import structure, large diffusion structure and large export structure; The large inlet structure is an inlet for liquid inlet over a large area; The topological microfluidic channel of the large diffusion structure is a three-section type, including: a front section, a middle section and a rear section; The front section includes a first axial fin and five first middle fins. The front section starts from the first axial fin. The first axial fin is arranged on the first symmetry axis. The five first middle fins are arranged in sequence and close to the second symmetry axis. Gaps are provided between the first axial fin and the five first middle fins. The fins of the front section introduce the inflowing cooling medium from the front section secondary channel to both sides, and the first middle fins cooperate with the front section main channel to bring the cooling medium close to the middle. The front section main channel gradually narrows. The middle section includes a second axial fin, a third axial fin, a fourth axial fin and three second middle fins. The middle section starts from the third axial fin. The third axial fin is arranged on the second symmetry axis. The three second middle fins are arranged in sequence and approach the first symmetry axis to reach the second axial fin on the first symmetry axis. At the same time, the fourth axial fin is arranged at the end position of the middle section. Gaps are provided between the second axial fin, the third axial fin, the fourth axial fin and the three second middle fins. The second middle fin of the middle section divides the cooling medium in the main channel of the front section into a first edge fluid. The first edge fluid gathers toward the middle under the influence of the fins of the middle section and is divided into a second edge fluid by the diverter fins of the middle section. The rear section includes three third middle fins, which are arranged in sequence from the second axial fins and close to the second symmetry axis, and gaps are provided between the three third middle fins; the second edge fluid is collected in the rear section, and the heat exchange area is increased by using the third middle fins of the rear section; The large outlet structure is an outlet for discharging liquid over a large area.

2. A design method for a scalable topology flow channel with high flux heat dissipation, the design method being used to design the scalable topology flow channel according to claim 1, It is characterized in that The following steps are involved: According to the actual working environment of the chip heat sink, a dual-plane pseudo three-dimensional topology optimization model is established, and the physical parameters, heat source distribution and cooling medium parameters of the chip heat sink are obtained; Determine the chip operating temperature, cooling fluid boundary conditions and heat source description form, determine the basic method of topology optimization problem based on the physical parameters of the chip heat sink, the heat source distribution and the cooling fluid parameters, and construct a topology optimization mathematical model by setting the topology optimization objective function; Performing finite element meshing on the three-dimensional topology optimization model and the topology optimization mathematical model, initializing model-related physical fields, calculating to obtain a topology optimization model, and performing parameter iteration on the topology optimization model to obtain a final topology optimization model; Based on the final topology optimization model, a spline function is used to fit a topology optimization two-dimensional model, and the topology optimization two-dimensional model is symmetrically expanded to obtain the expandable topology flow channel.

3. According to the design method of a scalable topology flow channel for high-throughput heat dissipation as described in claim 2, It is characterized in that The physical parameters include: radiator width W = 0.5 ~ 3mm, radiator length L 2 =5~20mm, radiator thickness z=0.1~1mm, radiator inlet length L in =0.5~2mm, radiator outlet length L out =0.5~2mm, radiator material, radiator material thermal conductivity k s , Radiator material constant pressure specific heat capacity C p,s and heat sink material density ρ s ; The heat source distribution includes: the heat source position is α 1 plane, the direction of application is from α 1 Plane conduction vertically to α 2 flat; The cooling medium parameters include: cooling medium composition, cooling medium heat conduction coefficient k f , cooling medium constant pressure specific heat capacity C p,f and cooling medium density ρ f .

4. According to the design method of a scalable topology flow channel for high-flux heat dissipation as described in claim 3, It is characterized in that The cooling medium boundary conditions include: inlet pressure p in =10~200Pa, outlet pressure p out =0Pa and inlet temperature T in =293.15~313.15K; The heat source description form is: F=h e (1-γ)(T s -T) Where F represents the total heat transfer in the design domain, h e Represents α 1 The heat transfer coefficient of the flat fin, γ represents the topology optimization design variable, T s represents the chip operating temperature, and T represents the temperature in the design domain.

5. According to the design method of a scalable topology flow channel for high-throughput heat dissipation as described in claim 4, It is characterized in that The method for constructing the topology optimization mathematical model includes: Based on the physical parameters of the chip heat sink, the distribution of the heat source and the parameters of the cooling medium, the variable density method is used to add laminar flow and fluid heat transfer physical fields to the pseudo three-dimensional topology optimization model, and the properties of the virtual solid-liquid mixed state in the design domain are defined by the material interpolation equation model: Among them, α(γ) represents the resistance coefficient, α f represents the reverse permeability in the fluid domain, α s represents the reverse permeability in the solid domain, ρ(γ) represents the virtual solid-liquid state density in the domain, C p (γ) represents the heat capacity of the virtual solid-liquid state in the domain, k(γ) represents the density thermal conductivity of the virtual solid-liquid state in the domain, q, q ρ ,qC p and q k represents the optimization parameter, μ in represents the inlet fluid viscosity, Da represents the Darcy number; Setting the total heat transfer F in the design domain as the topology optimization objective function; Set the constrained solid phase fraction and build the topology optimization mathematical model: Where Ψ represents the topology optimization design domain, V ψ Design domain volume for topology optimization, V s represents the set liquid phase fraction, p represents the pressure, u represents the velocity of the fluid in the x direction, v represents the velocity of the fluid in the y direction, and μ represents the dynamic viscosity of the fluid.

6. According to the design method of a scalable topology flow channel for high-throughput heat dissipation according to claim 5, It is characterized in that The method for obtaining the final topology optimization model includes: Finite element meshing is performed on the three-dimensional topology optimization model and the topology optimization mathematical model, and the resulting mesh independence and checkerboard problems are solved using a Helmholtz filter: in, represents the auxiliary function after the topology optimization design variable γ is filtered by the Helmholtz filter, represents the Hamiltonian operator, r m Indicates the filtering parameters. Indicates that The function after hyperbolic tangent projection, γ β represents the optimization parameters; Using the optimization iterative algorithm, by optimizing the parameters q, γ β and Da perform continuous processing on the filter function, set the maximum number of iteration steps and the iteration termination condition of each step, optimize the topology optimization objective function, and initialize the density field, temperature field and velocity field of the model to obtain the topology optimization model; The topology optimization model is used as the initial value, and the optimization parameters are set to solve the next topology optimization model, until the final convergence requirement is reached to complete the entire continuous optimization process, and the final topology optimization model is output.

7. According to the design method of a scalable topology flow channel for high-flux heat dissipation as described in claim 5, It is characterized in that The method for obtaining the expandable topological flow channel includes: The final topology optimization model is fitted using a B-spline function to obtain a fitted topology optimization two-dimensional model; The topology optimization two-dimensional model uses the bilateral symmetry setting in the boundary conditions, takes the bilateral symmetric boundaries as the symmetry axis, repeats the symmetric expansion multiple times, deletes the overlapping original symmetric condition boundaries, converts the remaining symmetric boundary conditions into internal boundary conditions, and obtains the expandable topology flow channel.

Images