Design method of finned liquid cooling plate

Abstract

The application relates to a design method of a fin liquid cooling plate, which comprises the following steps: processing a preset base plate, the upper end of the base plate is provided with an inlet channel and an outlet channel, at least one internal inlet is arranged in the inlet channel to form a distribution chamber, at least two internal outlets are arranged in the outlet channel to form a collection chamber; a fin array is fixedly arranged at the lower end of the base plate, the fin array comprises a plurality of parallel fins; a slit plate is arranged between the base plate and the fin array, a plurality of unequal-width slits are arranged on the slit plate, and the vertical projection of the internal inlet on the plane of the slit plate completely falls on the slit plate but not on the slits. According to the scheme, the flow distribution among the parallel fin flow channels of the high-density fin liquid cooling plate is as uniform as possible, the flow resistance of the liquid cooling plate is greatly reduced, the heat transfer is strengthened, the temperature uniformity of the core area of a chip is enhanced, the flow resistance is obviously reduced, and the mechanical energy consumption of the fluid is reduced.

Description

Technical Field

[0001] This invention relates to the technical field of finned heat sinks, specifically to a design method for a finned liquid cooling plate. Background Technology

[0002] High-density finned liquid coolers are core heat dissipation components in ultra-high heat flux density applications (chips, high-power IGBTs, electric drive systems / batteries in new energy vehicles, HPC clusters, etc.). Flow uniformity refers to the evenness of coolant flow distribution among the parallel finned channels. Flow uniformity design allows for full utilization of the heat exchange area of ​​all finned channels and the full release of convective heat transfer potential, avoiding heat exchange imbalances such as "excessive flow in some channels and insufficient flow in others." Especially for high heat flux density areas such as the core area of ​​chips, it can enhance the temperature uniformity of the cooler, prevent localized junction temperature exceeding limits, and avoid device frequency reduction, performance degradation, or even instantaneous failure. A cooler with excellent flow uniformity performance can reduce the flow resistance of the liquid cooler, reduce the power consumption of the circulating pump, and directly reduce the system's PUE (Power Usage Effectiveness). Flow uniformity can also meet heat dissipation requirements at lower flow rates, reducing coolant consumption and system maintenance costs. It can be said that flow uniformity is a prerequisite for high-density fins to realize their heat exchange potential and has become one of the core bottleneck technologies for liquid cooling. On the other hand, increasing the flow rate can further improve the heat dissipation power of the chip, but it also leads to a significant increase in the flow resistance of the liquid cooler and pumping energy consumption. Therefore, innovation is urgently needed in the design of low-flow-resistance liquid coolers that can adapt to high flow rates.

[0003] However, existing high-density finned liquid cooling plates (commonly known as "toothed plates") mostly adopt a liquid inlet in the middle and liquid outlet on both sides of the fin array, or a liquid inlet on both sides and liquid outlet in the middle of the fin array, or even a liquid inlet at one end and an outlet at the other, due to considerations such as structural simplicity, ease of processing, or insufficient design optimization. These flow design methods all suffer from uneven flow distribution between parallel fin channels and high flow resistance due to the long distance the fluid travels between the fins. Summary of the Invention

[0004] To at least address one of the shortcomings of the existing technology, the present invention aims to provide a design method for a finned liquid cooling plate.

[0005] The technical solution adopted by this invention to solve its problem is: a design method for a finned liquid cooling plate, comprising:

[0006] The substrate board has an inlet channel and an outlet channel at its upper end. The inlet channel has at least one internal inlet to form at least one liquid distribution chamber, and the outlet channel has at least two internal outlets to form at least two liquid collection chambers.

[0007] A fin array comprising multiple parallel fins, the fin array being fixedly disposed at the lower end of the substrate plate;

[0008] A slotted plate is provided between the substrate plate and the fin array. The slotted plate has several slots, and the vertical projection of the internal liquid inlet onto the plane of the slotted plate falls entirely on the slotted plate rather than on the slots.

[0009] Furthermore, specifically, the liquid distribution chamber and its internal liquid inlet are symmetrically arranged about the center line of the fin array, and the liquid collection chamber and its internal liquid outlet are symmetrically arranged about the center line of the fin array. The center line of symmetry refers to any center line with respect to the length or width direction of the fin array.

[0010] Furthermore, specifically, the slits corresponding to the internal liquid inlet are set as unequal width slits, and the smaller the distance between them and the internal liquid inlet buckle, the narrower the width; the slits corresponding to the internal liquid outlet are set as equal width slits.

[0011] Furthermore, specifically, the flow channel of the internal liquid outlet is inclined to the plane of the slotted plate.

[0012] Furthermore, specifically, the internal liquid inlet is positioned perpendicular to the plane of the slotted plate.

[0013] Furthermore, specifically, the closer the internal inlet is to the inlet of the liquid inlet channel, the larger the cross-sectional area of ​​the internal inlet; the farther the internal inlet is from the inlet of the liquid inlet channel, the smaller the cross-sectional area of ​​the internal inlet.

[0014] Furthermore, specifically, the optimal number of dispensing chambers and collecting chambers is determined through a preset first optimization algorithm, including:

[0015] First, determine the quantitative relationship between the separating chamber and the collecting chamber;

[0016] Establish geometric models of finned liquid cooling plates with different numbers of dispensing and collecting chambers;

[0017] Obtain inlet and outlet pressure drop and maximum temperature data of heating surface under arbitrary liquid cooling plate geometric model;

[0018] Under the constraint that the maximum temperature of the heating surface does not exceed a preset threshold, the number of liquid distribution chambers corresponding to the scheme with the minimum pressure drop at the inlet and outlet of the finned liquid cooling plate is selected as the optimal value, and the width of the corresponding individual liquid chamber is determined at the same time.

[0019] Furthermore, specifically, based on the optimal number of dispensing chambers and collecting chambers, a pre-set second optimization algorithm is used to determine the design of unequal-width slits, including:

[0020] The arbitrarily selected seam is divided into three sections: core section, intermediate transition section, and two side sections, and a parameterized width distribution model is established.

[0021] Perform multi-objective computational fluid dynamics simulation optimization within the set parameter range, while considering two optimization objectives: minimizing flow resistance and ensuring uniform flow distribution along the slot length.

[0022] Under the constraint that the maximum temperature of the heating surface does not exceed the allowable threshold, the optimal distribution of unequal width slots is determined by a preset optimization algorithm.

[0023] Furthermore, specifically, the preset optimization algorithm is Pareto front analysis or weighted summation method.

[0024] The beneficial effects of this invention are:

[0025] (1) Significantly reduced flow resistance: The multi-parallel flow path design with multiple liquid distribution chambers and multiple liquid collection chambers shortens the flow path of fluid entering and exiting the fin array and effectively reduces flow resistance; combined with the unequal width slot optimization design, local resistance loss is further reduced while ensuring the uniformity of flow distribution.

[0026] (2) Uniform flow distribution: By symmetrical arrangement of the liquid distribution chamber and the design principle of the narrowest core section of the unequal width slot, the coolant is ensured that it will not directly jet and impact a local area of ​​the fin array, but will be evenly distributed to the gaps between each parallel fin after being buffered by the slot plate.

[0027] (3) Uniform temperature field distribution: The optimized design makes the temperature distribution of the heating surface more uniform, effectively avoiding local junction temperature exceeding the standard, which is conducive to ensuring the safe operation of high heat flux density devices such as chips.

[0028] (4) The heat exchange capacity remains good: Although the single flow rate is reduced by multiple flow paths, the simulation verification shows that the overall heat dissipation capacity of the cold plate can still meet the requirements, and the average temperature of the heating surface is controlled within a reasonable range. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly explained below. Obviously, the described drawings are only a part of the embodiments of the present invention, and not all of them. Those skilled in the art can obtain other design schemes and drawings based on these drawings without creative effort.

[0030] Figure 1 (a) and Figure 1 (b) are two perspectives of the overall structure of the finned liquid cooling plate design method proposed in this invention;

[0031] Figure 2This is a schematic diagram of the internal flow channel of a finned liquid cooling plate design method proposed in this invention;

[0032] Figure 3 This is a cross-sectional structural schematic diagram of a design method for a finned liquid cooling plate proposed in this invention.

[0033] Figure 4 This is a schematic diagram showing the velocity distribution of the symmetrical cross section in the middle of the liquid cooling plate during the simulation process of the design method of the finned liquid cooling plate proposed in this invention.

[0034] Figure 5(a) and Figure 5(b) are schematic diagrams of the velocity distribution between the fin array of the finned liquid cooling plate proposed in this invention and the fin array of the finned liquid cooling plate of a typical existing design, respectively.

[0035] Figure 6(a) and Figure 6(b) are schematic diagrams of the temperature field of the heat dissipation surface of the finned liquid cooling plate proposed in this invention and the finned liquid cooling plate of a typical existing design, respectively.

[0036] Figure 7 A comparison diagram of the flow rates of three inlet branches during the simulation process of the design method of the finned liquid cooling plate proposed in this invention;

[0037] Figure 8(a) and Figure 8(b) are schematic diagrams of the pressure drop of the finned liquid cooling plate proposed in this invention and the finned liquid cooling plate of a typical existing design, respectively.

[0038] Figure 9 This is a structural schematic diagram of another cross section of the design method for a finned liquid cooling plate proposed in this invention;

[0039] Figure 10 This is a perspective view of the internal structure of a finned liquid cooling plate proposed in this invention. Detailed Implementation

[0040] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention. Furthermore, all connection relationships mentioned herein do not simply refer to direct connection of components, but rather to the ability to form a better circuit structure by adding or reducing connecting components according to specific implementation conditions. The various technical features in this invention can be combined interactively without contradicting each other.

[0041] Reference Figure 1 (a) Figure 1 (b) Figure 2 , Figure 3 as well as Figure 9 Example 1: A design method for a finned liquid cooling plate, applicable to heat transfer situations involving phase change. When two phases exchange heat, the separating chamber is a liquid, and the collecting chamber is a gas-liquid mixture. The method includes:

[0042] The substrate board 100 has a liquid inlet channel 110 and a liquid outlet channel 120 at its upper end. The liquid inlet channel 110 has at least one internal liquid inlet 111, which forms at least one liquid distribution chamber 112. The liquid outlet channel 120 has at least two internal liquid outlets 121, which forms at least two liquid collection chambers 122. (The number of internal liquid outlets 121 is equal to the number of internal liquid inlets 111 + 1.)

[0043] The fin array 200 includes a plurality of parallel fins, and the fin array 200 is fixedly disposed at the lower end of the substrate plate 100;

[0044] A slotted plate 300 is provided between the substrate plate 100 and the fin array 200. The slotted plate 300 has a plurality of slots, and the vertical projection of the internal liquid inlet 111 onto the plane of the slotted plate 300 falls entirely on the slotted plate 300 rather than on the slots.

[0045] In this embodiment 1, a design is adopted with multiple liquid distribution chambers 112 and their internal liquid inlets 111, and multiple liquid collection chambers 122 and their multiple internal liquid outlets 121. This effectively shortens the flow path of fluid entering and exiting the fin array 200, and the parallel flow of multiple processes reduces the flow velocity, thereby significantly reducing flow resistance. Under high flow rates, this design can also keep the flow resistance of the entire liquid cooling plate within a reasonable range.

[0046] The liquid flow direction of this finned liquid cooling plate is as follows: the liquid cooling plate coolant flows in through the liquid inlet channel 110, flows through the internal liquid inlet 111 into the liquid distribution chamber 112, the coolant in the liquid distribution chamber 112 flows to the slotted plate 300 for buffering, and then enters the fin array 200 through the slots. After being acted upon by the fin array 200, it returns to the liquid collection chamber 122 through the slots, and then flows through the internal liquid outlet 121 to the liquid outlet channel 120, completing one complete cycle.

[0047] Furthermore, when the liquid coolant passes through the inlet channel 110 and flows into the distribution chamber 112 through the internal inlet 111, it does not directly impact the fins. Instead, it is buffered and blocked by the slotted plate 300, converting dynamic pressure into static pressure. This prevents most of the internal inlet jet from directly impacting the fin array 200. The distribution chamber 112 also slows down the flow, preventing high-velocity liquid from directly impacting a specific area of ​​the fin array 200. This avoids uneven flow distribution, where a larger flow distribution occurs in the fin area and a smaller flow distribution in other areas. Simultaneously, the reduced flow velocity is converted into pressure potential energy, resulting in a more uniform pressure distribution in the distribution chamber 112, further promoting uniform flow distribution in the fin array 200.

[0048] In a preferred embodiment of the present invention, the liquid distribution chamber 112 and its internal liquid inlet 111 are symmetrically arranged about the center line of the fin array 200, and the liquid collection chamber 122 and its internal liquid outlet 121 are symmetrically arranged about the center line of the fin array 200. The center line of symmetry refers to any center line in terms of the length or width of the fin array.

[0049] In this preferred embodiment, the above design ensures that the flow field has good symmetry about the fin array 200, which is beneficial to the uniformity of the flow field. The liquid inlet channel and the internal liquid inlet are arranged on the center line to better distribute the liquid evenly. The internal liquid outlet is also arranged on the center line to make the overall flow symmetrical about the center line, so that the flow is more uniform.

[0050] In a preferred embodiment of the present invention, the slit corresponding to the internal liquid inlet 111 is designated as the first slit 311 and is configured as a slit of unequal width, with the width being narrower as the distance between it and the internal liquid inlet buckle is smaller. The slit corresponding to the internal liquid outlet 121 is designated as the second slit 312 and is configured as a slit of equal width.

[0051] In this preferred embodiment, the core section closest to the internal inlet 111 is the narrowest, with a width approximately 0.5-0.65 times the width of the two sides, and a length approximately 2-3 times the maximum length of the internal inlet 111. The width of the middle section is approximately 0.7-0.75 times the width of the two sides, and the length on one side is approximately 0.7-1.0 times the length of the core section. The maximum width of the slit on both sides is designed to be 1mm-8mm depending on the number of dispensing chambers 112 (a smaller value is used when there are many dispensing chambers 112, and a larger value is used when there are few dispensing chambers 112). The specific width can be slightly adjusted according to the actual situation. Simulation analysis shows that the design of the equal-width outlet slit does not affect the uniformity of the process. Therefore, adopting the equal-width outlet slit design can reduce the complexity of processing, thereby reducing processing costs.

[0052] Reference Figure 10In a preferred embodiment of the present invention, the flow channel of the internal liquid outlet 121 (i.e., the flow channel formed by the internal liquid outlet) is inclined to the plane of the slotted plate 300.

[0053] In this preferred embodiment, this method ensures that when the cold plate outflow channel deviates from the symmetrical center line of the fin array 200, the internal liquid outlet 121 can still be arranged on the symmetrical center line of the fin array 200, which is beneficial to the symmetry and uniformity of the flow field distribution. In addition, the inclined setting of the water outlet hole is also convenient for processing. That is, the inclined arrangement of the flow channel of the internal liquid outlet is for the rationality of the arrangement, ensuring that the flow channel of the internal liquid outlet must avoid the liquid inlet of the liquid inlet channel and the flow channel of the internal liquid inlet.

[0054] In a preferred embodiment of the present invention, the internal liquid inlet 111 is specifically arranged perpendicular to the plane of the slotted plate 300.

[0055] As a preferred embodiment of the present invention, specifically, the liquid inlet of the liquid inlet channel 110 (refer to...) Figure 10 The closer the internal inlet 111 is, the larger its cross-sectional area; the further away from the inlet of the inlet channel 110 (refer to...). Figure 10 The further away the internal liquid inlet 111 is, the smaller its cross-sectional area becomes.

[0056] In this preferred embodiment, based on the basic principles of fluid dynamics, this method ensures that the flow rate of each internal inlet 111 is substantially the same.

[0057] In a preferred embodiment of the present invention, specifically, the optimal number of dispensing chambers and collecting chambers is determined by a preset first optimization algorithm, including...

[0058] First, determine the quantitative relationship between the separating chamber and the collecting chamber;

[0059] Establish geometric models of finned liquid cooling plates with different numbers of dispensing and collecting chambers;

[0060] Obtain inlet and outlet pressure drop and maximum temperature data of heating surface under arbitrary liquid cooling plate geometric model;

[0061] Under the constraint that the maximum temperature of the heating surface does not exceed a preset threshold, the number of liquid distribution chambers corresponding to the scheme with the minimum pressure drop at the inlet and outlet of the finned liquid cooling plate is selected as the optimal value, and the width of the corresponding individual liquid chamber is determined at the same time.

[0062] Specifically, based on the optimal number of dispensing chambers and collecting chambers, a pre-set second optimization algorithm is used to determine the design of unequal-width slits, including:

[0063] The arbitrarily selected seam is divided into three sections: core section, intermediate transition section, and two side sections, and a parameterized width distribution model is established.

[0064] Perform multi-objective computational fluid dynamics simulation optimization within the set parameter range, while considering two optimization objectives: minimizing flow resistance and ensuring uniform flow distribution along the slot length.

[0065] Under the constraint that the maximum temperature of the heating surface does not exceed the allowable threshold, the optimal distribution of unequal width slots is determined by a preset optimization algorithm.

[0066] In this preferred embodiment, the basic idea of ​​the above two-step optimization design method is as follows:

[0067] The first step is to determine the optimal number of liquid chambers:

[0068] (1) Establish multiple sets of liquid cooling plate geometric models with different numbers of liquid distribution chambers, and arrange the liquid distribution chambers and liquid collection chambers symmetrically about the symmetrical center line of the fin array;

[0069] (2) Perform computational fluid dynamics fluid-structure interaction heat transfer simulation analysis on each model to obtain the pressure drop at the inlet and outlet of the cold plate and the highest temperature of the heating surface;

[0070] (3) Under the constraint that the maximum temperature of the heating surface does not exceed the allowable threshold, the number of liquid distribution chambers corresponding to the scheme with the smallest pressure drop at the inlet and outlet of the cold plate is selected as the optimal value, and the width of the corresponding single liquid chamber is determined at the same time.

[0071] The second step is to optimize the design of the unequal width slots based on the determined optimal number of liquid chambers:

[0072] (1) Divide the first seam into three sections: core section, intermediate transition section, and two side sections, and establish a parameterized width distribution model;

[0073] (2) Perform multi-objective computational fluid dynamics simulation optimization within the set parameter range, while considering two optimization objectives: minimizing flow resistance and ensuring uniform flow distribution along the length of the slot.

[0074] (3) Under the constraint that the maximum temperature of the heating surface does not exceed the allowable threshold, the optimal distribution of unequal width slots is determined by Pareto front analysis or weighted summation method.

[0075] The specific process is as follows:

[0076] By using computational fluid dynamics simulation, the number of liquid chambers and the width of the slots are optimized in a coordinated manner to minimize the flow resistance of the liquid cooler while meeting the chip's heat dissipation requirements.

[0077] Step 1: Optimization Design Method for the Number of Liquid Chambers

[0078] (a) Determine the configuration relationship between the inlet chamber and the outlet chamber

[0079] (1) The number of liquid collection chambers Equal to the number of the dispensing chambers ( Add 1, that is:

[0080] (1)

[0081] (2) Ensure that the length of the fin array through which the fluid flows from each inlet chamber to the collection chamber is basically equal, thereby ensuring the balance of flow resistance and the uniformity of flow distribution.

[0082] (II) The correlation mechanism between the number of analytical liquid chambers and flow resistance

[0083] (1) The more liquid inlet chambers there are, the shorter the average distance the fluid travels from the liquid inlet channel through the internal liquid inlet into the liquid distribution chamber, and then through the slots into the fin array, and the smaller the corresponding friction loss.

[0084] Specifically, when a parallel flow path design with multiple liquid distribution chambers and multiple liquid collection chambers is adopted, the fluid flow path is effectively shortened, and the parallel flow of multiple processes reduces the flow velocity, thereby significantly reducing flow resistance.

[0085] (2) At the same time, the increase in the number of flow paths leads to a decrease in the flow rate of each flow path, a decrease in the average flow velocity between the fin arrays, and a reduction in the convective heat transfer capacity.

[0086] Therefore, there exists an optimal number of liquid chambers that minimizes flow resistance while ensuring heat dissipation.

[0087] (III) Setting optimization objectives and constraints

[0088] (1) Optimization objective: When the chip operates at its highest power At that time, while meeting the highest temperature of the chip's heating surface Not exceeding the specified threshold Under these circumstances, the pressure drop at the inlet and outlet of the cold plate It should be as small as possible.

[0089] (2) Temperature constraint: The maximum temperature of the heating surface must not exceed the maximum allowable temperature limit to ensure the safe operation of the chip.

[0090] (2)

[0091] (3) Geometric constraints: Width of all dispensing and collecting chambers The sum should be less than or equal to the total width of the heat dissipation surface. Add 10mm. This constraint ensures that the overall structural dimensions of the liquid cooling plate meet the actual assembly space requirements.

[0092] (3)

[0093] (iv) Determine the optimal number of liquid chambers through computational fluid dynamics simulation.

[0094] The specific steps are as follows:

[0095] (1) Initialize the search range: Based on the geometric constraints, deduce the maximum possible number of liquid inlet chambers. Set the initial search value to 2.

[0096] (2) Number of inlet chambers for each candidate Perform the following loop operation:

[0097] ① Establish the geometric model of the liquid cooling plate:

[0098] The pre-designed finned plate is fabricated according to the aforementioned design method;

[0099] The upper end of the finned plate is provided with the liquid inlet channel and the liquid outlet channel;

[0100] The liquid inlet channel is provided with multiple internal liquid inlets and connected to multiple liquid distribution chambers, the number of which is equal to the current candidate value. ;

[0101] The liquid outlet channel is equipped with multiple internal liquid outlets connected to multiple liquid collection chambers, the number of which is equal to the candidate value plus one. +1;

[0102] Each liquid distribution chamber and its internal liquid inlet are arranged symmetrically about the symmetrical center line of the fin array;

[0103] Each liquid collection chamber and its internal liquid outlet are arranged symmetrically about the symmetrical center line of the fin array.

[0104] ② Set the initial parameters for the slot (for the first simulation):

[0105] The slotted plate has several slots, which are temporarily set to be slots of equal width.

[0106] The width of the equal-width slit can be initially determined based on the number of separation chambers. When there are many separation chambers, a smaller value (such as 0.5mm to 1.5mm) should be used, and when there are few separation chambers, a larger value (such as 1.5mm to 2.5mm) should be used.

[0107] ③ Set simulation boundary conditions:

[0108] Apply power to the heating surface (i.e., the chip mounting surface) at the chip's maximum operating power. ;

[0109] Each inlet is set with a design flow rate. and the temperature of the imported liquid ;

[0110] Back pressure conditions are set at the outlet of the liquid outlet channel.

[0111] ④ Perform computational fluid dynamics (CFD) fluid-structure interaction heat transfer simulation analysis: use multiple computing graphics cards (GPGPU) to accelerate numerical simulation calculations, and perform post-processing and result analysis under typical working conditions.

[0112] ⑤ Obtain simulation result data: Record the pressure drop values ​​at the inlet and outlet of the cold plate. Record the highest temperature value of the heating surface. Record auxiliary parameters such as the average temperature of the heating surface and the uniformity of velocity distribution between fin arrays.

[0113] ⑥ Determine if the temperature constraint is met: If > If the number of candidate inlet chambers does not meet the design requirements, skip to the next candidate value; if If it is, then it is recorded as a valid design scheme, and the corresponding ( , , , Data set.

[0114] (3) Select from all valid design schemes The minimum number of inlet chambers is the optimal solution: the number of inlet chambers corresponding to this solution is the optimal number of inlet chambers. The width of a single liquid chamber is the optimal liquid chamber width. .

[0115] Step 2: Optimization Design Method for Unequal Width Seams

[0116] (a) Definition of technical features

[0117] 1. The first seam

[0118] The slots corresponding to the internal liquid inlet are channels for coolant to flow from the distribution chamber into the fin array.

[0119] 2. The second seam

[0120] The slots corresponding to the internal liquid outlets are channels through which coolant flows back to the collection chamber from the fin array.

[0121] 3. Core Section

[0122] The narrowest part of the first slit, closest to the internal liquid inlet.

[0123] 4. Intermediate transition section

[0124] The first seam is located in the transitional section between the core segment and the two side segments, where the width gradually changes.

[0125] 5. Both sides

[0126] The widest edge portion of the first slit, furthest from the internal liquid inlet.

[0127] 6. Local flow distribution function

[0128] The instantaneous flow rate variation of the coolant at various positions along the length of the first slit.

[0129] (II) Specific Steps of the Design Method

[0130] Given a fixed number of optimal liquid chambers, the design of unequal-width slots is optimized using the following method:

[0131] 1. Set optimization goals and design principles

[0132] (1) Bi-objective optimization:

[0133] The primary objective is to ensure a uniform coolant flow distribution along the length of the first slot.

[0134] The second objective is to reduce the flow resistance of the entire heat dissipation plate (i.e., the pressure drop at the inlet and outlet of the heat dissipation plate). (Minimum)

[0135] (2) Core design principles:

[0136] The faster the flow rate, the narrower the width of the first slit should be, so as to achieve a more uniform redistribution of flow rate; the smaller the distance between the first slit and the internal inlet, the narrower the width of the first slit should be, so as to reduce the impact of the direct jet on the local area of ​​the fin array.

[0137] 2. Determine the geometric constraints and parameter range of unequal width slots.

[0138] (1) Minimum width constraint of the first seam: the width of the core segment It should be smaller than the optimal single liquid chamber width. Subtract the diameter of the internal inlet , Right now:

[0139] (4)

[0140] This constraint ensures that the coolant does not directly impact the fin array, but is instead blocked and buffered by the slotted plate, converting dynamic pressure into static pressure.

[0141] (2) Width and proportion range of each segment:

[0142] The width range of the two sides is designed to be greater than or equal to 0.5 mm and less than or equal to 10 mm, depending on the number of dispensing chambers. Specifically, a smaller value (e.g., 0.5 mm to 3 mm) is used when there are many dispensing chambers, and a larger value (e.g., 3 mm to 10 mm) is used when there are few dispensing chambers.

[0143] The width of the core segment Approximately the width of the two side sections 0.4 to 0.65 times, that is:

[0144] (5)

[0145] The width of the intermediate transition section Approximately the width of the two side sections 0.6 to 0.8 times, that is:

[0146] (6)

[0147] (3) Range of length proportions for each segment:

[0148] The length of the core segment Approximately the diameter of the internal inlet. 2 to 5 times, that is:

[0149] (7)

[0150] The single-sided length of the intermediate transition section Approximately the length of the core segment 0.5 to 1.1 times, that is:

[0151] (8)

[0152] After the lengths of the core segment and the middle segment are determined, the lengths of the two side segments are... Naturally determined.

[0153] Establish a parameterized model for unequal width strip seams

[0154] (1) Segmentation definition: Divide the first seam into three segments along its length:

[0155] The first section is the core section: located near the internal liquid inlet, it is the narrowest in width and is used to reduce the direct impact of high-speed jets on the fins.

[0156] The second section is the intermediate transition section: located between the core section and the two side sections, with a gradually changing width, used for smooth traffic transition and distribution;

[0157] The third section consists of two side sections: located on the side furthest from the internal liquid inlet, with the widest width, used to ensure that coolant enters the gaps between each parallel fin evenly.

[0158] In practice, more segments can be used to achieve better flow distribution, depending on the actual situation, but at least three segments are required, which also takes into account the ease of processing.

[0159] (2) Definition of parameter vector:

[0160] Define the optimization parameter vector =[ , , , , ],in:

[0161] Width of both sides, in millimeters;

[0162] : The width ratio of the core segment to the two side segments, dimensionless;

[0163] The ratio of the width of the middle transition section to the width of the two side sections, dimensionless;

[0164] Core section length and internal inlet diameter The ratio, dimensionless;

[0165] : The ratio of the length of one side of the intermediate transition section to the length of the core section, dimensionless.

[0166] (3) Establish the seam width distribution function Based on the parameter vector Generate a width variation curve along the length of the first seam:

[0167] (9)

[0168] In the formula, ; .

[0169] According to the parameter vector Set the length distribution of each segment of the first seam:

[0170] (10)

[0171] 4. Perform parameter scanning for multi-objective optimization

[0172] (1) Initialize the search grid within the set parameter range:

[0173] Value range: Determine the initial value based on the number of liquid chambers, such as 1 to 3 mm;

[0174] Range of values: see equation (5);

[0175] Range of values: see equation (6);

[0176] Range of values: see equation (7);

[0177] Range of values: see equation (8).

[0178] (2) For each candidate parameter combination Perform the following loop operation:

[0179] ① Generate the geometric model of unequal width slots:

[0180] Call the optimal liquid chamber configuration determined in the first step ( , )

[0181] According to the parameter vector Set the width and length distribution of each segment of the first seam.

[0182] ② Establish a complete geometric model of the liquid cooling plate and perform CFD simulation. Boundary conditions are set, including setting the total flow rate to the design flow rate value. The initial condition is that the inlet flow rate of each dispensing chamber is evenly distributed; the power applied to the heating surface is the chip's maximum operating power. wait.

[0183] ③ Obtain simulation result data, including: the pressure drop values ​​at the inlet and outlet of the cold plate. The highest temperature value of the heating surface The average temperature value of the heating surface Local flow distribution along the length of the first slit .

[0184] ④ Calculate the objective function value:

[0185] First target component Characterizing the effect of flow resistance optimization:

[0186]

[0187] Second target component Characterizing the optimization effect of flow distribution uniformity along the length of the slot:

[0188]

[0189] in This represents the average flow rate.

[0190] The comprehensive evaluation indicators are calculated using a weighted summation method:

[0191]

[0192] The weighting coefficients α and β are determined according to the design focus (e.g., when the emphasis is on reducing resistance, α=0.6 and β=0.4 can be taken).

[0193] ⑤ Determine if the constraints are satisfied:

[0194] like > If so, record the parameter combination as an invalid scheme; If the design is valid, it is recorded as a valid design scheme. The comprehensive evaluation index is then calculated and recorded as a valid scheme.

[0195] (3) Determine the optimal parameter combination using a multi-objective optimization decision-making method:

[0196] Option 1: Pareto front analysis. (Draw) Scatter plots are used to identify the set of non-dominated solutions, from which the parameter combination that satisfies the temperature constraint and has the most balanced overall performance is selected as the optimal solution. .

[0197] Option 2: Optimization after weighted summation. After setting weight coefficients, directly select the parameter combination with the smallest comprehensive evaluation index as the optimal solution. .

[0198] 5. Output and Verification

[0199] (1) Output the optimal result:

[0200] The optimal width of both sides ;

[0201] The optimal core segment ratio ;

[0202] The optimal intermediate transition ratio ;

[0203] The corresponding optimal width distribution function of the first slit .

[0204] (2) Final verification simulation: A complete geometric model of the liquid-cooled plate is established using the optimal parameter combination, and a final CFD simulation is performed to verify: A: < (Superior to typical design schemes); B: ≤ (Meets temperature constraints).

[0205] (3) Iterative adjustment (if necessary): If the highest temperature of the heating surface still exceeds the allowable threshold in the final verification, the parameter range is appropriately relaxed or the width of the two sides is increased, and step 4 is re-executed until convergence meets the temperature constraint requirements.

[0206] Furthermore, the following simulation verification was performed on this scheme:

[0207] Step 1: Design a liquid cooling plate according to the method determined in this invention.

[0208] Step 2: Perform computational fluid dynamics (CFD) simulation analysis of fluid-structure interaction heat transfer to verify the effectiveness of the design method.

[0209] The geometric model of the liquid-cooled plate designed by the method of this invention is imported into CFD simulation software for preprocessing such as geometric model processing, mesh generation, boundary condition definition, and physical model setting.

[0210] Multiple computing graphics cards (GPGPUs) are used to accelerate numerical simulation calculations.

[0211] Typical working condition simulation post-processing and result analysis

[0212] velocity field

[0213] Figure 4 The velocity distribution at the symmetrical cross-section of the liquid-cooled plate reflects the velocity distribution within the intermediate liquid inlet branch, its distribution chamber 112, and the flow channels between the fins. Figure 4 As can be seen, the fluid diffuses along the liquid distribution chamber 112 to both sides of the entire fin array 200, and the velocity distribution between the fins is basically uniform.

[0214] Figure 5(a) shows the velocity distribution between the fin arrays 200 of the present invention, and Figure 5(b) shows the velocity distribution between arrays in a typical existing design. From the comparison of the velocity field distribution between the fin arrays 200 in Figure 5, it can be seen that the flow field distribution between the entire fin array 200 of the present invention is very uniform; the velocity uniformity between arrays in a typical existing design is relatively poor, with the flow rate in the middle section being significantly higher than in other sections, leading to heat exchange imbalance in the liquid cooling plate and affecting its temperature uniformity. CFD simulations demonstrate that the flow field uniformity effect of the method of the present invention is very good.

[0215] (2) Temperature field

[0216] Simulations were performed under the conditions of 1800W power at the heating surface and a flow rate of 3.77L / min. Figure 6(a) shows the temperature field of the cold plate heat dissipation surface, and Figure 6(b) shows the temperature field of a typical liquid-cooled plate heat dissipation surface. From the temperature field comparison in Figure 6, it can be seen that the temperature field uniformity of the cold plate heat dissipation surface of the present invention is significantly better than that of a typical liquid-cooled plate heat dissipation surface. The average temperature of the cold plate heat dissipation surface of the present invention is 56.93℃, while the average temperature of a typical liquid-cooled plate heat dissipation surface is 56.2℃, with a difference of less than 1℃. This indicates that despite the reduced flow rate, the present invention still has good heat exchange capacity, and the average temperature control of the cold plate is excellent. The highest temperature of the cold plate heat dissipation surface of the present invention is 63.0℃, which is 1.5℃ lower than the highest temperature of 64.5℃ of a typical liquid-cooled plate heat dissipation surface. This demonstrates that the present invention ensures good heat exchange performance and effectively reduces the highest temperature of the heat dissipation surface, further reducing it from the maximum temperature limit of the heat dissipation surface in actual engineering, thus contributing to chip safety. CFD simulations show that the flow field uniformity of the present invention leads to good temperature field uniformity and simultaneously provides good heat dissipation capacity.

[0217] (3) Comparison of flow rates of the three inlet branches

[0218] Figure 7 This section compares the flow rates of the three internal inlet branches (blue line for Mass Flow Monitor A, yellow line for Mass Flow Monitor B, and red line for Mass Flow Monitor C) when the total flow rate is 3.77 L / min. The flow rates of each branch are as follows:

[0219] The flow rates are 0.0213 kg / s, 0.0208 kg / s, and 0.0199 kg / s. The maximum absolute difference in flow rate among the inlet branches is 0.0014 kg / s, and the relative difference is 6.57%. This demonstrates that the method of gradually reducing the flow area of ​​each inlet branch starting from the liquid-cooled plate inlet can make the flow rates of each branch very similar, which is beneficial for achieving uniform flow distribution throughout the heat exchange area.

[0220] (4) Pressure drop

[0221] Figure 8(a) shows the pressure drop of the liquid cooling plate of the present invention, and Figure 8(b) shows the pressure drop of a typical conventional liquid cooling plate. Figure 7 The pressure drop comparison of the liquid cooling plates shows that, with a flow rate of 3.77 L / min and the same fin array 200 structure and size, the pressure drop of the cold plate of this invention is 6410.7 Pa, which is much lower than the pressure drop of 11092.4 Pa of a typical liquid cooling plate, a reduction of 42.2%! While achieving better heat exchange effect, the significant reduction in the flow resistance of the liquid cooling plate is conducive to reducing pumping energy consumption, thereby achieving energy saving.

[0222] 4. Comparison of flow resistance under variable flow rate

[0223] Table 1 compares the pressure drop of the liquid cooling plate of the present invention with that of a typical existing liquid cooling plate at various flow rates. As can be seen from Table 1, the pressure drop of the liquid cooling plate of the present invention is significantly lower than that of a typical existing liquid cooling plate at different flow rates; and even at a maximum flow rate of 6.033 L / min, the pressure drop of the liquid cooling plate of the present invention is still well controlled, decreasing by 9205 Pa compared to a typical existing liquid cooling plate, a reduction of 37.4%!

[0224] Table 1 Comparison of flow resistance under varying flow rates

[0225]

[0226] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A design method for a finned liquid cooling plate, characterized in that, include: A preset substrate board is obtained through processing. The upper end of the substrate board is provided with a liquid inlet channel and a liquid outlet channel. The liquid inlet channel is provided with at least one internal liquid inlet to form at least one liquid distribution chamber. The liquid outlet channel is provided with at least two internal liquid outlets to form at least two liquid collection chambers. A fin array is fixedly disposed at the lower end of the substrate plate, the fin array comprising a plurality of parallel fins; A slotted plate is provided between the substrate plate and the fin array. The slotted plate has a number of slots of unequal width, and the vertical projection of the internal liquid inlet onto the plane of the slotted plate falls entirely on the slotted plate rather than on the slots. The optimal number of dispensing chambers and collecting chambers is determined by a preset first optimization algorithm. The design of unequal width slits is determined by a preset second optimization algorithm based on the optimal number of dispensing chambers and collecting chambers.

2. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the liquid distribution chamber and its internal liquid inlet are symmetrically arranged about the center line of the fin array, and the liquid collection chamber and its internal liquid outlet are symmetrically arranged about the center line of the fin array. The center line of symmetry refers to any center line relative to the longitudinal or transverse direction of the fin array.

3. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the slits corresponding to the internal liquid inlet are set as unequal width slits, and the smaller the distance between them and the internal liquid inlet, the narrower the width; the slits corresponding to the internal liquid outlet are set as equal width slits.

4. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the flow channel of the internal liquid outlet is inclined to the plane of the slotted plate.

5. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the internal liquid inlet is positioned perpendicular to the plane of the slotted plate.

6. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the closer the internal inlet is to the inlet of the liquid inlet channel, the larger its cross-sectional area; the farther the internal inlet is from the inlet of the liquid inlet channel, the smaller its cross-sectional area.

7. The design method of a finned liquid cooling plate according to claim 1, characterized in that, Specifically, the optimal number of dispensing chambers and collecting chambers is determined through a preset first optimization algorithm, including: First, determine the quantitative relationship between the separating chamber and the collecting chamber; Establish geometric models of finned liquid cooling plates with different numbers of liquid distribution chambers and liquid collection chambers; Obtain inlet and outlet pressure drop and maximum temperature data of heating surface under arbitrary liquid cooling plate geometric model; Under the constraint that the maximum temperature of the heating surface does not exceed a preset threshold, the number of liquid distribution chambers corresponding to the scheme with the minimum pressure drop at the inlet and outlet of the finned liquid cooling plate is selected as the optimal value, and the width of the corresponding individual liquid chamber is determined at the same time.

8. The design method of a finned liquid cooling plate according to claim 7, characterized in that, Specifically, based on the optimal number of dispensing chambers and collecting chambers, a pre-set second optimization algorithm is used to determine the design of unequal-width slits, including: The arbitrarily selected seam is divided into three sections: core section, intermediate transition section, and two side sections, and a parameterized width distribution model is established. Perform multi-objective computational fluid dynamics simulation optimization within the set parameter range, while considering two optimization objectives: minimizing flow resistance and ensuring uniform flow distribution along the slot length. Under the constraint that the maximum temperature of the heating surface does not exceed the allowable threshold, the optimal distribution of unequal width slots is determined by a preset optimization algorithm.

9. The design method of a finned liquid cooling plate according to claim 8, characterized in that, Specifically, the preset optimization algorithm is Pareto front analysis or weighted summation method.

Images