Method and application of designing tree-shaped heat pipe structure using particle swarm evolution

Abstract

The present invention discloses a method and application for designing the structure of a tree-shaped heat pipe using particle swarm evolution. The present invention designs the evaporation section of the tree-shaped heat pipe by geometric modeling of the tree-shaped heat pipe, establishing the objective function for designing the tree-shaped heat pipe according to the requirements of thermal design, and calculating the objective function using a particle swarm evolution algorithm, thereby obtaining a tree-shaped heat pipe structure with good heat transfer characteristics. The structure can improve the heat dissipation capacity of the entire heat pipe in some areas with large heat flux density by increasing the distribution density of microchannels in these areas, and has a broad research prospect.

Description

Technical Field

[0001] The invention belongs to the field of heat dissipation equipment or heat exchange equipment, and particularly relates to a method for designing the structure of a tree-shaped heat pipe by using particle swarm evolution and its application. Background Art

[0002] In recent years, the application of electronic products and electronic equipment has become more and more extensive, and some electronic components and integrated circuits are usually used in their manufacture. With the continuous improvement of circuit design and manufacturing process level, and the gradual entry into nanotechnology, the number of transistors on integrated circuit chips has increased exponentially, which has improved the integration of devices and the operating speed. This has also brought some other problems that need to be solved. For example, due to the increase in device integration, its heat generation and heat flux density have also increased dramatically. Therefore, heat treatment has become an important factor that needs to be considered in the design. Efficiently transferring the heat generated by the device to the surrounding environment to reduce the operating temperature of the device is an important aspect of ensuring the reliability of electronic devices and related electronic products and electronic equipment.

[0003] Conventional heat transfer mainly includes three methods: convection, radiation and conduction. For some electronic products, since they generate a lot of heat when working, if conventional heat transfer methods are used, the heat dissipation effect is poor. For this, it is necessary to use enhanced heat transfer methods, such as pulsating heat pipes.

[0004] Pulsating heat pipes are generally made of capillaries with a diameter of about 0.5mm to 5mm that are bent in a serpentine shape. Structurally, pulsating heat pipes can be divided into three parts: evaporation section, insulation section and condensation section. When performing heat transfer and heat dissipation, the pulsating heat pipe places the evaporation section in the heat receiving area and the condensation section in the heat dissipation area. For pulsating heat pipes, under a vacuum state, the working fluid is injected into its pipeline at a certain liquid filling ratio, and a series of liquid plugs and bubbles are formed under the action of surface tension. During operation, the working fluid absorbs energy and generates bubbles in the evaporation section, and then the bubbles gradually merge with the gas plug, causing the internal pressure of the gas plug to increase, and pushing the gas plug and liquid plug to move toward the condensation section; at the same time, the gas plug in the condensation section shrinks or ruptures due to cooling, thereby forming an unbalanced vapor pressure field between the evaporation section and the condensation section, and driving the working fluid to flow on the pipeline.

[0005] The working fluid in the pulsating heat pipe is also affected by gravity, capillary force, gas-liquid two-phase pressure difference and viscous resistance when moving, which will lead to a more complex flow pattern, making the alternating distribution of liquid and bubbles in the working fluid during the flow process more random, thereby causing changes in the temperature, pressure and flow velocity of the working fluid, so that the flow of the working fluid has the characteristics of phase change heat transfer and oscillation heat transfer, which improves the efficiency of the entire heat transfer to a certain extent.

[0006] The structure of the pulsating heat pipe can be divided into open and closed types according to the connection method of its internal pipes. The open pulsating heat pipe adopts a separate sealing method at both ends, and the working fluid cannot circulate in the pipe. The closed pulsating heat pipe connects the two ends of the capillary tube to form a complete serpentine loop, which is conducive to the circulation of the working fluid, and therefore has better heat transfer performance than the open pulsating heat pipe.

[0007] The evaporation section of the pulsating heat pipe is placed in the heat receiving area, and its structure has an important influence on the heat transfer performance of the entire heat pipe. How to effectively design the structure of the pulsating heat pipe and obtain a pulsating heat pipe with good heat transfer performance is still a research direction. Summary of the invention

[0008] The primary purpose of the present invention is to overcome the deficiencies of the prior art and provide a method for designing the structure of a tree-shaped heat pipe using particle swarm evolution.

[0009] Another object of the present invention is to provide an application of a method for designing the structure of a tree-shaped heat pipe using particle swarm evolution.

[0010] The purpose of the present invention is achieved through the following technical solutions:

[0011] A method for designing a tree-shaped heat pipe structure using particle swarm evolution comprises the following steps:

[0012] Perform geometric modeling of the tree-shaped heat pipe, that is, determine the initial shape of the geometric structure of the tree-shaped heat pipe;

[0013] According to the requirements of thermal design, an objective function for optimizing the design of tree-shaped heat pipes is established;

[0014] The particle swarm evolutionary algorithm is used to calculate the objective function and obtain a tree-shaped heat pipe structure with better heat transfer characteristics.

[0015] The design of the tree-shaped heat pipe structure is preferably to design the structure of the evaporation section of the pulsating heat pipe in a tree-shaped structure.

[0016] The tree-shaped heat pipe structure is designed. For the condensation section of the pulsating heat pipe, the same tree-shaped structure as the evaporation section can be used, or a conventional microchannel structure can be used.

[0017] The tree-like structure can adjust the distribution of microchannel branches according to the distribution of heat flux density in the heat source area; in some areas with large heat flux density, the heat dissipation capacity of the entire heat pipe in these areas can be improved by increasing the distribution density of microchannels in these areas, and the temperature gradient distribution of the heat source area after heat dissipation can be made more uniform.

[0018] The geometric modeling of the tree-shaped heat pipe is preferably to perform geometric modeling on the structure of the evaporation section of the pulsating heat pipe, and more preferably to perform modeling on the number of microchannel branches of the evaporation section of the pulsating heat pipe, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch.

[0019] The objective function is defined as F(b):

[0020]

[0021] The optimization design is to make the value of F(b) reach a larger value, and the corresponding individual tree-shaped structure heat pipe has a smaller thermal resistance, thereby having a better heat transfer performance.

[0022] The method of calculating the objective function using the particle swarm evolution algorithm comprises the following steps:

[0023] (1) Encode the individuals. An individual represents a structure of the evaporation section, which contains information such as the number of microchannel branches, the length of each branch, the cross-sectional area of ​​each branch, and the location of each branch. The values ​​of this information are combined into a whole as the encoding of an individual;

[0024] (2) For a given individual b, use and solve the corresponding heat balance equations to obtain the temperature value at each node position in the steady state, thereby calculating the thermal resistance R of the heat pipe b ; Define the following objective function F(b):

[0025]

[0026] When the value of F(b) reaches a larger value, the tree-shaped heat pipe has better heat transfer performance; here, the fitness value (suitability) of an individual b is defined as F(b);

[0027] (3) A population is formed by a plurality of individuals; a particle swarm evolution algorithm is used to find a better individual through the evolution of the population; the better individual corresponds to a structure of an evaporation section, and the heat pipe corresponding to the better individual has a lower thermal resistance and better heat transfer performance;

[0028] Let s represent the evolutionary generation of the population, P(s) represent the s-th generation population, and L represent the size of the population P(s), i.e., the number of individuals in the population. Here, L is a positive integer given in advance. X i represents the i-th individual in the population P(s), where X i (i=1,2,…,L).

[0029] The steps of the particle swarm evolution algorithm described in step (3) are preferably as follows:

[0030] S1. Set parameter s = 0; calculate the fitness value of each individual in the population P(s);

[0031] S2. For each individual in P(s), perform particle swarm evolution operation so that one parent individual produces one child individual;

[0032] S3. All the sub-individuals generated in step 2 form a sub-population Q(s); calculate the fitness value of each individual in Q(s);

[0033] S4. Perform the selection operation, that is, select L individuals from the parent population P(s) and the child population Q(s); set: s←s+1; the new population P(s+1) consists of these L individuals;

[0034] S5. If the value of s is less than a predetermined value, go to step S2, otherwise go to step S6;

[0035] S6. Select the individual with the best fitness value in the population P(s+1) and use it as the output of the entire algorithm.

[0036] The particle swarm evolution operation described in step S2 is preferably implemented as follows:

[0037] Each individual in P(s) is regarded as a particle; first, for each particle, if its fitness value is better than the best fitness value in history (named pBest), the current value is set as the new pBest; second, the particle with the best fitness value among all particles is selected as gBest; third, for each particle, the particle velocity is calculated according to expression (1), and the particle position is updated according to expression (2).

[0038] v ij (s+1)=w·v ij (s)+c1·r1·(z ij (s)-x ij (s))+c2·r2·(z gj (s)-x ij (s)) (1)

[0039] x ij (s+1)=x ij (s)+v ij (s+1) (2)

[0040] Here, each individual in the population P(s) is regarded as a particle (a point) in a multidimensional space (here we assume it is a D-dimensional space). The i-th particle is represented by X i =(x il ,x i2 ,…,x iD). For the i-th particle, use Z i =(z il ,z i2 ,…,z iD ) represents the best position before it (i.e., the position that gives the best fitness value). g represents the best individual in the population (the individual corresponding to the best fitness value). V i represents the running speed of particle i (reflected in the change of position), V i =(v il ,v i2 ,…,v iD ).

[0041] The velocity vector and position vector are updated according to expressions (1) and (2), respectively, where 1≤i≤L, 1≤j≤D. Here i is the number of the particle, j represents the jth dimension of the particle, and s represents the evolutionary generation of the algorithm. Parameters c1 and c2 are two positive constants, r1 and r2 are two random numbers in the range [0,1]. Parameter w is the weight, and its value is a positive constant.

[0042] The selection operation described in step S4 is preferably implemented as follows:

[0043] For the parent population P(s) and the child population Q(s), let is the union of P(s) and Q(s), that is The selection operation is to determine which individuals will be retained to the next generation and which individuals will be Removed. Each individual X in i Must compete with other individuals for the chance to be preserved to the next generation.

[0044] right Each individual X in i , use expression (3) to assign it a weight U i .

[0045]

[0046]

[0047] Where q is a positive integer (e.g. q=8), which means that individual X i will and from q other individuals X randomly selected from j (j=1,2,…,q) for comparison. Parameter λ j The value of is 0 or 1, defined by expression (4); F i and F j They are individual X i and Xj The fitness value of τ is a uniformly distributed random number in the range of [0,1].

[0048] For an individual X i ,from Randomly select q individuals X from j (j=1,2,…,q), and each individual X j With individual X i After comparison, we can calculate the individual X i The weight is U i When The weights U of all individuals in i After all the calculations are done, All individuals in are sorted by the weight U i The value of individual X is in descending order i Arrange them in order. Select the first L individuals and put them into the population P(s+1) as the parent individuals of the next generation.

[0049] The predetermined value in step S5 is a predetermined maximum number of iterations of the algorithm.

[0050] The fitness value described in step S6 is the best individual, and the structure of the tree-shaped heat pipe in the evaporation section corresponding to the individual is the desired evaporation section structure, which has a smaller thermal resistance and therefore has a better heat transfer performance.

[0051] The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution can be applied in the preparation of a pulsating heat pipe.

[0052] Principle of the present invention:

[0053] The structure of the evaporation section of a conventional pulsating heat pipe is generally a serpentine loop formed by bending a capillary tube with a smaller diameter in a serpentine shape. For example, the capillary tubes are arranged vertically and bent at the bottom and top to form a serpentine loop, such as Figure 2 As shown. Because there are some bends, these bends will have a certain impact on the flow and circulation of the working medium of the heat pipe. In this regard, the capillaries (microchannels) can be arranged in a tree shape to make a tree-shaped heat pipe, such as Figure 3 shown.

[0054] The structural design of the tree-shaped heat pipe in the evaporation section mainly includes the following four aspects: the number of microchannel branches, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch. Here, the range of the evaporation section can be regarded as a rectangular area on the plane, and the position of a microchannel branch is the position of the branch in this rectangular area. The above four aspects play a decisive role in the heat transfer performance of the tree-shaped heat pipe. The present invention designs the evaporation section with a tree structure by using a particle swarm evolution algorithm to obtain a tree-shaped evaporation section structure with better heat transfer performance.

[0055] In order to calculate the average temperature of the evaporation section, it is necessary to obtain the temperature values ​​at some positions of the evaporation section. This is because by calculating the average value of the temperature values ​​at these positions, the average temperature of the evaporation section can be obtained. For this purpose, the evaporation section is placed on a rectangular area, the rectangular area is divided into equal grids, and all the microchannel branches of the evaporation section are divided into three-dimensional structures to obtain a large number of three-dimensional solid units. In general, all microchannel branches are composed of these three-dimensional solid units. This three-dimensional solid unit is called a microelement here.

[0056] Since the rectangular area is divided into equal grids, each grid point or node (i, j) in the rectangular area may have a corresponding microelement placed (when there is a microchannel branch on the node), or may not have a corresponding microelement placed (when there is no microchannel branch on the node). Each microelement will generate a certain amount of heat, part of which will be transferred to the adjacent microelement, and the other part will be carried away by the air in the environment through convection. After a certain period of time, thermal equilibrium will be reached between this heat generation and heat conduction, forming a steady-state heat distribution. The heat balance equation in the laws of thermodynamics is used below to solve the temperature of each node in the evaporation section.

[0057] The basic law of heat conduction, Fourier's law, is as follows:

[0058]

[0059] Where Q is the heat flux, which represents the amount of heat passing through a given area per unit time, and the unit is W; is the temperature gradient, in ℃ / m; A is the thermal conductivity area, in m 2 ; k is the thermal conductivity of the material, the unit is W / (m·℃).

[0060] By using Fourier's law, we can obtain the differential equation for heat conduction. Take a cubic microelement such as Figure 4 As shown, the three sides of the microelement are dx, dy, and dz respectively. According to the basic law of heat conduction, the heat flowing into the left side of the microelement along the positive direction of the x-axis is:

[0061]

[0062] The heat flowing out from the right side of the microelement is:

[0063]

[0064] Therefore, in the x direction, the net heat inflow caused by heat conduction is:

[0065]

[0066] Similarly, we can write two equations for the y and z directions that are similar to the x direction:

[0067]

[0068] Therefore, the total net heat inflow dQ should be equal to the heat stored in the parallelepiped:

[0069]

[0070] If heat Q0 occurs in each unit time and unit space in the microelement, the heat generated in the microelement is:

[0071] dQ′=Q0dxdydz;

[0072] The heat retained by the heat conducting surface in the microelement is added to the heat generated in the microelement to increase the internal energy of the microelement. Therefore, the rate of increase of internal energy can be expressed as:

[0073]

[0074] The above formula shows the amount of heat required when the temperature of a microelement rises by a certain degree. P is the specific heat capacity (J / kg℃), ρ is the density (kg / m 3 ), is the time rate of change of temperature, and dxdydz is the volume of the microelement.

[0075] According to the law of conservation of energy, the rate of increase of heat storage should be equal to the sum of the net heat flowing into the microelement by heat conduction and the heat generated in the microelement: dQ+dQ′=dE, that is:

[0076]

[0077] If the thermal conductivity is stable, The above formula can be written as:

[0078]

[0079] If there is no internal heat source in the heat conduction system, for stable heat conduction, we have:

[0080]

[0081] The above formula is the differential equation expression for three-dimensional stable heat conduction.

[0082] Within a given period of time, each microelement in the rectangular area where the evaporation section is located can reach a stable temperature. The energy balance principle of the microelement is used to derive and obtain the heat balance equation of each node.

[0083] For a node (i, j), assuming that the third dimension perpendicular to the rectangular area is a unit length, the heat introduced from node (i+1, j) to node (i, j) per unit time can be expressed as:

[0084]

[0085] Here T i+1,j and T i,j are the temperatures of nodes (i+1,j) and (i,j) respectively; σ = 1 / ρ, parameter ρ is the heat transfer coefficient between nodes; parameter ξ is the ratio of the mesh size in the horizontal and vertical directions. If ε is the mesh size in the horizontal axis and η is the mesh size in the vertical axis, then ξ = ε / η. Here (σ·ξ) is the thermal resistance between nodes (i+1,j) and (i,j). Similarly, the heat expression introduced into node (i,j) by nodes (i-1,j), (i,j+1), and (i,j-1) in unit time can be obtained.

[0086] For the evaporation section, if the third dimension perpendicular to the rectangular area is used for convective heat transfer, the heat introduced into the node (i, j) perpendicular to the rectangular area can be expressed as:

[0087] M0=(T0-T i,j )·θ0·η;

[0088] Where T0 is the air temperature of the external environment, θ0 is the heat transfer coefficient between the air and the node (i, j) in the direction perpendicular to the rectangular area. Node (i, j) is a heat source node, and it will also generate a certain amount of heat per unit time, which is recorded as H i,j In a stable state, the following expression can be obtained according to the energy balance principle:

[0089] M i+1,j +M i-1,j +M i,j+1 +M i,j-1 +H i,j +M0=0;

[0090] Right now

[0091]

[0092] The above equation is the heat balance equation for the internal nodes. For the boundary nodes, their corresponding heat balance equations can be derived similarly.

[0093] Using the above analysis of various nodes, the thermal balance equations of each microelement on each microchannel branch in the entire evaporation section are listed separately to form a linear equation group. By solving the equation group, the temperature value at each node position in the steady state can be obtained.

[0094] The structural design of the tree-shaped heat pipe in the evaporation section mainly includes the following four aspects: the number of microchannel branches, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch. For different structures, the corresponding thermal conductivity and thermal resistance are different. The particle swarm evolution algorithm is used to find the structure with smaller thermal resistance.

[0095] First, encode the individuals. An individual represents a structure of the evaporation section, which contains four types of information: the number of microchannel branches, the length of each branch, the cross-sectional area of ​​each branch, and the location of each branch. The values ​​of these four types of information are combined into a whole as the encoding of an individual.

[0096] For a given individual b, use and solve the corresponding heat balance equations to obtain the temperature value at each node position in the steady state, and thus calculate the thermal resistance R of the heat pipe. b . Define the following objective function F(b):

[0097]

[0098] The tree-shaped heat pipe with a larger value of F(b) has better heat transfer performance. Here, the fitness value (suitability) of an individual b is defined as F(b).

[0099] Secondly, a population is formed by multiple individuals. Through the evolution of the population, a better individual is found. The better individual here is a structure of an evaporation section, and the corresponding heat pipe has a lower thermal resistance and better heat transfer performance.

[0100] Let s represent the evolutionary generation of the population, P(s) represent the s-th generation population, and L represent the size of the population P(s), i.e., the number of individuals in the population. Here, L is a positive integer given in advance. X i represents the i-th individual in the population P(s), where X i (i=1,2,…,L).

[0101] A particle swarm evolution algorithm for finding better individuals is given below, and its calculation steps are as follows.

[0102] First, encode the individuals and then calculate the fitness value of each individual. Secondly, use the following Algorithm 1 to generate a new population.

[0103] Algorithm 1

[0104] Step 1: Set parameter s=0; calculate the fitness value of each individual in the population P(s).

[0105] Step 2: For each individual in P(s), perform particle swarm evolution operation so that one parent individual produces one child individual.

[0106] Step 3: All the sub-individuals generated in step 2 form a sub-population Q(s). Calculate the fitness value of each individual in Q(s).

[0107] Step 4: Perform the selection operation, that is, select L individuals from the parent population P(s) and the child population Q(s). Set: s←s+1. The new population P(s+1) consists of these L individuals.

[0108] Step 5: If the value of s is less than a predetermined value, go to step 2, otherwise go to step 6.

[0109] Step 6: Select the individual with the best fitness value in the population P(s+1) and use it as the output of the entire algorithm.

[0110] The detailed implementation of Algorithm 1 is described below. For step 2 of Algorithm 1, its specific implementation is as follows: each individual in P(s) is regarded as a particle. First, for each particle, if its fitness value is better than the best fitness value in history (named pBest), the current value is set as the new pBest. Second, the particle with the best fitness value among all particles is selected as gBest. Third, for each particle, the particle velocity is calculated according to expression (1), and the particle position is updated according to expression (2).

[0111] v ij (s+1)=w·v ij (s)+c1·r1·(z ij (s)-x ij (s))+c2·r2·(z gj (s)-x ij (s)) (1)

[0112] x ij (s+1)=x ij (s)+v ij (s+1) (2)

[0113] Here, each individual in the population P(s) is regarded as a particle (a point) in a multidimensional space (here we assume it is a D-dimensional space). The i-th particle is represented by X i =(x il ,x i2 ,…,x iD ). For the i-th particle, use Z i =(z il ,z i2 ,…,z iD ) represents the best position before it (i.e., the position that gives the best fitness value). g represents the best individual in the population (the individual corresponding to the best fitness value). V i represents the running speed of particle i (reflected in the change of position), V i =(v il ,v i2 ,…,v iD ).

[0114] The velocity vector and position vector are updated according to expressions (1) and (2), respectively, where 1≤i≤L, 1≤j≤D. Here i is the number of the particle, j represents the jth dimension of the particle, and s represents the evolutionary generation of the algorithm. Parameters c1 and c2 are two positive constants, r1 and r2 are two random numbers in the range [0,1]. Parameter w is the weight, and its value is a positive constant.

[0115] For step 4 of Algorithm 1, the selection operation is implemented as follows: let the parent population be P(s) and the child population be Q(s), and let is the union of P(s) and Q(s), that is The selection operation is to determine which individuals will be retained to the next generation and which individuals will be Removed. Each individual X in i Must compete with other individuals for the chance to be preserved to the next generation.

[0116] right Each individual X in i , use expression (3) to assign it a weight U i .

[0117]

[0118]

[0119] Where q is a positive integer (e.g. q=8), which means that individual X i will and from q other individuals X randomly selected fromj (j=1,2,…,q) for comparison. Parameter λ j The value of is 0 or 1, defined by expression (4); F i and F j They are individual X i and X j The fitness value of τ is a uniformly distributed random number in the range of [0,1].

[0120] For an individual X i ,from Randomly select q individuals X from j (j=1,2,…,q), and each individual X j With individual X i After comparison, we can calculate the individual X i The weight is U i When The weights U of all individuals in i After all the calculations are done, All individuals in are sorted by the weight U i The value of individual X is in descending order i Arrange them in order. Select the first L individuals and put them into the population P(s+1) as the parent individuals of the next generation.

[0121] For step 5 of algorithm 1, if the value of s is less than a predetermined value, go to step 2, where the predetermined value is a predetermined maximum number of iterations of the algorithm, for example, its value is 600.

[0122] By running Algorithm 1, an individual with a better fitness value can be obtained. The structure of the tree-shaped heat pipe in the evaporation section corresponding to the individual is the desired evaporation section structure, which has a smaller thermal resistance and therefore has better heat transfer performance.

[0123] Compared with the prior art, the present invention has the following advantages and effects:

[0124] The present invention adopts a tree structure for the evaporation section of the pulsating heat pipe, and for the condensation section of the pulsating heat pipe, the tree structure of the present invention, which is the same as the evaporation section, can be used, or a conventional microchannel structure can be used. This tree structure allows the distribution of microchannel branches to be adjusted according to the distribution of the heat flux density in the heat source area. In some areas with large heat flux density, the heat dissipation capacity of the entire heat pipe in these areas can be improved by increasing the distribution density of the microchannels in these areas, and the distribution of the temperature gradient of the heat source area after heat dissipation can be made more uniform. In addition, compared with conventional pulsating heat pipes, the start-up time and start-up temperature of the pulsating heat pipe using the tree structure, as well as the average wall temperature of the evaporation section are lower, so it can better adapt to a variety of complex heat source conditions and has broad research prospects. When designing a pulsating heat pipe using a tree structure, a key issue is the design of the microchannel branches, which mainly includes the following four aspects: the number of microchannel branches, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch, etc. These four aspects have a decisive effect on the heat transfer performance of the heat pipe. The present invention designs the evaporation section of the pulsating heat pipe using a tree structure by using a particle swarm evolution algorithm, thereby obtaining the evaporation section structure of the pulsating heat pipe with good heat transfer performance, which can be widely used in practice. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] Figure 1 It is a flow chart of designing the evaporation section structure of the tree-shaped pulsating heat pipe according to the present invention.

[0126] Figure 2 It is a structural block diagram of a pulsating heat pipe.

[0127] Figure 3 It is a structural schematic diagram of a tree-shaped heat pipe.

[0128] Figure 4 It is a structural diagram of a microelement.

[0129] Figure 5 It is a schematic structural diagram obtained by designing the evaporation section of the tree-shaped heat pipe according to the present invention. DETAILED DESCRIPTION

[0130] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0131] The present invention provides a design method using particle swarm evolution for the evaporation section structure of a tree-shaped heat pipe, and its implementation steps are as follows: first, geometric modeling of the tree-shaped heat pipe is performed, that is, the initial shape of the geometric structure of the tree-shaped heat pipe is determined; second, according to the requirements of thermal design, an objective function for optimizing the design of the tree-shaped heat pipe is established; third, the objective function is calculated using a particle swarm evolution algorithm, that is, Algorithm 1, to obtain a tree-shaped heat pipe structure with better heat transfer characteristics. The specific implementation steps are as follows:

[0132] (1) For the evaporation section of the pulsating heat pipe, the present invention uses a tree structure; for the condensation section of the pulsating heat pipe, the tree structure of the present invention can be used, or a conventional microchannel structure can be used.

[0133] (2) The structural design of the tree-shaped heat pipe in the evaporation section mainly includes the following four aspects: the number of microchannel branches, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch. For different evaporation section structures, the thermal conductivity and thermal resistance of the corresponding pulsating heat pipe are different.

[0134] (3) Use the particle swarm evolution algorithm to find the evaporation section structure with smaller thermal resistance.

[0135] First, encode the individuals. An individual represents a structure of the evaporation section, which contains four types of information: the number of microchannel branches, the length of each branch, the cross-sectional area of ​​each branch, and the location of each branch. The values ​​of these four types of information are combined into a whole as the encoding of an individual.

[0136] For a given individual b, use and solve the corresponding heat balance equations to obtain the temperature value at each node position in the steady state, and thus calculate the thermal resistance R of the heat pipe. b The fitness value of individual b is defined as:

[0137]

[0138] Secondly, a population is formed by multiple individuals. The optimal individual is obtained through the evolution of the population, i.e., using Algorithm 1. The optimal individual here is a structure of an evaporation section, which corresponds to a heat pipe with a lower thermal resistance and better heat transfer performance.

[0139] The following takes the tree structure design of the evaporation section of a pulsating heat pipe as an example to illustrate the specific implementation steps of the present invention.

[0140] Example: According to the given design specification requirements, realize the structural design of a pulsating heat pipe. The requirements given in the design specification are: when the working fluid is deionized water, the filling rate is 70%, the entire flow channel is distributed on a square aluminum plate of 80mm×80mm, the microchannel matrix material is pure aluminum, and the power of the heat source is 5W, the thermal resistance of the designed pulsating heat pipe is lower than 1.2K / W, and the total volume of all branches is less than 130mm 3 .

[0141] In the implementation, the present invention adopts a tree structure to design the evaporation section of the pulsating heat pipe, and adopts a columnar structure, i.e., a structure similar to a rectangular parallelepiped, for each branch in the tree structure; for the condensation section of the pulsating heat pipe, it also adopts the same structure as the evaporation section. For the design of the evaporation section of the pulsating heat pipe, the geometric modeling of the tree-shaped heat pipe is firstly carried out, and then the objective function for the optimization design of the tree-shaped heat pipe is established, and then the particle swarm evolution algorithm, i.e., Algorithm 1, is used to calculate the objective function. Here, the scale L of the population P(s) is 20. For the parameters c1 and c2 in expression (1), and w is the weight, their values ​​are: c1=0.45, c2=0.55, and w=0.6 respectively. For the parameter q in expression (3), its value is q=8. In Algorithm 1, the maximum number of iterations is set to 100, that is, when the number of iterations of the algorithm is greater than or equal to 100, the operation is terminated.

[0142] Algorithm 1 is set using the above parameters. By running Algorithm 1, the design result of the evaporation section of the pulsating heat pipe is obtained as follows: the tree structure of the evaporation section contains six branches a1, a2, b1, b2, a3, and b3, as shown in Figure 5 As shown; the thickness of each branch is 0.5mm; the length and width of branches a1 and a2 are 74mm and 0.7mm respectively; the length and width of branches b1 and b2 are 48mm and 0.5mm respectively; the length and width of branches a3 and b3 are 36mm and 0.4mm respectively; the total volume of all six branches is 90.2mm 3 ;The thermal resistance of the entire pulsating heat pipe is 0.83K / W.

[0143] The result of running Algorithm 1 is that the thermal resistance of the entire pulsating heat pipe is 0.83K / W, which is less than the 1.2K / W required by the design specification. The total volume of the six branches is 90.2mm 3 Smaller than the 130mm required by the design specification 3 .

[0144] Therefore, the designed pulsating heat pipe meets the given design specification requirements. By using the tree-shaped structure of the evaporation section of the pulsating heat pipe obtained by the method provided by the present invention, the distribution of the microchannel branches can be adjusted according to the distribution of the heat flux density in the heat source area. In some areas with large heat flux density, the heat conduction to the heat source can be better achieved by increasing the distribution density of the microchannels in these areas. Through the combined effect with the condensation section, the heat transfer performance of the entire pulsating heat pipe can be improved.

[0145] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for designing the structure of a tree-shaped heat pipe using particle swarm evolution, characterized in that The following steps are involved: Perform geometric modeling of the tree-shaped heat pipe, that is, determine the initial shape of the geometric structure of the tree-shaped heat pipe; According to the requirements of thermal design, an objective function for optimizing the design of tree-shaped heat pipes is established; The objective function is calculated using a particle swarm evolutionary algorithm to obtain a tree-shaped heat pipe structure; The geometric modeling is to model the number of microchannel branches of the evaporation section of the heat pipe, the length of each branch, the size of the cross-sectional area of ​​each branch, and the position of each branch; The steps of using the particle swarm evolution algorithm to calculate the objective function are as follows: (1) Encoding the individuals; an individual represents a structure of the evaporation section, which includes information such as the number of microchannel branches, the length of each branch, the cross-sectional area of ​​each branch, and the position of each branch; the values ​​of this information are combined into a whole as the encoding of an individual; (2) For a given individual b, use and solve the corresponding heat balance equations to obtain the temperature value at each node position in the steady state, thereby calculating the thermal resistance R of the heat pipe b ; Define the following objective function F(b): When the value of F(b) reaches a larger value, the tree-shaped heat pipe has better heat transfer performance; here, the fitness value (suitability) of an individual b is defined as F(b); (3) A population is formed by a plurality of individuals; a particle swarm evolution algorithm is used to obtain a better individual through the evolution of the population; the better individual corresponds to a structure of an evaporation section, and the heat pipe corresponding to the better individual has a lower thermal resistance and better heat transfer performance; Let s represent the evolutionary generation of the population, P(s) represent the s-th generation population, and L represent the size of the population P(s), i.e., the number of individuals in the population. Here, L is a positive integer given in advance. X i represents the i-th individual in the population P(s), where X i (i=1,2,···,L).

2. The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution according to claim 1, characterized in that The particle swarm evolution algorithm described in step (3) includes the following steps: first, each individual in the parent population generates a child individual, and then all the generated child individuals constitute a child population; second, from the union of the parent population and the child population, determine which individuals will enter the next generation by using selection operations and individual competition.

3. The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution according to claim 1, characterized in that: The particle swarm evolution algorithm in step (3) includes the following steps: S1. Set parameter s = 0; calculate the fitness value of each individual in the population P(s); S2. For each individual in P(s), perform particle swarm evolution operation so that one parent individual produces one child individual; S3. All the sub-individuals generated in step 2 form a sub-population Q(s); calculate the fitness value of each individual in Q(s); S4, perform the selection operation, that is, select L individuals from the parent population P(s) and the child population Q(s); Set: s←s+1; the new population P(s+1) consists of these L individuals; S5. If the value of s is less than a predetermined value, go to step S2, otherwise go to step S6; S6. Select the individual with the best fitness value in the population P(s+1) and use it as the output of the entire algorithm.

4. The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution according to claim 3, characterized in that: The particle swarm evolution operation described in step S2 includes the following steps: Consider each individual in P(s) as a particle; first, for each particle, if its fitness value is better than the best fitness value in history, set the current value as the new pBest; second, select the particle with the best fitness value among all particles as gBest; third, for each particle, calculate the particle speed according to expression (1), and update the particle position according to expression (2); v ij (s+1) = w·v ij (s) + c1·r1· (z ij (s)-x ij (s)) + c2·r2· (z gj (s)-x ij (s)) (1); x ij (s+1) = x ij (s) + v ij (s+1) (2); Here, each individual in the population P(s) is regarded as a particle in a multidimensional space; the i-th particle is represented as X i =(x il ,x i2 ,···,x iD );For the i-th particle, use Z i =(z il ,z i2 ,···,z iD ) represents the best position before it; g represents the best individual in the population; V i represents the running speed of particle i, V i =(v il ,v i2 ,···,v iD ); The velocity vector and position vector are updated according to expressions (1) and (2) respectively, where 1≤i≤L, 1≤j≤D; i is the particle number, j represents the jth dimension of the particle, and s represents the evolutionary generation of the algorithm; parameters c1 and c2 are two normal numbers, r1 and r2 are two random numbers in the range [0,1]; parameter w is the weight, and its value is a normal number.

5. The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution according to claim 3, characterized in that: The selection operation described in step S4 includes the following steps: For the parent population P(s) and the child population Q(s), let is the union of P(s) and Q(s), that is The selection operation is to determine which individuals will be retained to the next generation and which individuals will be Removed; Each individual X in i Must compete with other individuals for the chance to be preserved to the next generation; right Each individual X in i , use expression (3) to assign it a weight U i ; Where q is a positive integer, which means that individual X i will and from q other individuals X randomly selected from j (j=1,2,···,q) for comparison; parameter λ j The value of is 0 or 1, defined by expression (4); F i and F j They are individual X i and X j The fitness value of ; τ is a uniformly distributed random number in the range of [0,1]; For an individual X i ,from Randomly select q individuals X from j (j=1,2,···,q), and for each individual X j With individual X i After comparison, we can calculate the individual X i The weight is U i The value of The weights U of all individuals in i After all the calculations are done, All individuals in are sorted by the weight U i The value of individual X is in descending order i Arrange them in sequence; select the first L individuals and put them into the population P(s+1) as the parent individuals of the next generation.

6. The method for designing the structure of a tree-shaped heat pipe using particle swarm evolution according to claim 3, characterized in that: The predetermined value in step S5 is the predetermined maximum number of iterations of the algorithm; The fitness value described in step S6 is the best individual, and the structure of the tree-shaped heat pipe in the evaporation section corresponding to the individual is the desired evaporation section structure, which has a smaller thermal resistance and therefore has a better heat transfer performance.

7. Use of the method for designing the structure of a tree-shaped heat pipe using particle swarm evolution as claimed in any one of claims 1 to 6 in the preparation of a pulsating heat pipe.

Images