CN116796593A - Free modeling and optimal design method for horizontal wave-shaped channel - Google Patents
Free modeling and optimal design method for horizontal wave-shaped channel Download PDFInfo
- Publication number
- CN116796593A CN116796593A CN202310624959.1A CN202310624959A CN116796593A CN 116796593 A CN116796593 A CN 116796593A CN 202310624959 A CN202310624959 A CN 202310624959A CN 116796593 A CN116796593 A CN 116796593A
- Authority
- CN
- China
- Prior art keywords
- channel
- function
- cold plate
- horizontal wave
- modeling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013461 design Methods 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000005457 optimization Methods 0.000 claims abstract description 28
- 238000012546 transfer Methods 0.000 claims abstract description 22
- 230000017525 heat dissipation Effects 0.000 claims abstract description 17
- 238000010606 normalization Methods 0.000 claims abstract description 6
- 238000012937 correction Methods 0.000 claims abstract description 4
- 239000012530 fluid Substances 0.000 claims description 34
- 239000011159 matrix material Substances 0.000 claims description 12
- 239000007787 solid Substances 0.000 claims description 9
- 238000013519 translation Methods 0.000 claims description 9
- 239000007788 liquid Substances 0.000 claims description 7
- 238000010438 heat treatment Methods 0.000 claims description 6
- 230000014509 gene expression Effects 0.000 claims description 5
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims description 4
- 229910052782 aluminium Inorganic materials 0.000 claims description 4
- 239000000463 material Substances 0.000 claims description 4
- 239000011343 solid material Substances 0.000 claims description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000012887 quadratic function Methods 0.000 claims description 2
- 238000001816 cooling Methods 0.000 description 5
- 230000002068 genetic effect Effects 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 241000143252 Idaea infirmaria Species 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000004907 flux Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000012938 design process Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- WYTGDNHDOZPMIW-RCBQFDQVSA-N alstonine Natural products C1=CC2=C3C=CC=CC3=NC2=C2N1C[C@H]1[C@H](C)OC=C(C(=O)OC)[C@H]1C2 WYTGDNHDOZPMIW-RCBQFDQVSA-N 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000001627 detrimental effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Algebra (AREA)
- Computing Systems (AREA)
- Fluid Mechanics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a free modeling and optimizing design method of a horizontal wave-shaped channel, which adopts a cubic spline function to represent amplitude, a wavelet function to represent wavelength and a truncated function to represent waveform, and realizes free correction of the cross section shape and size of the channel through a rational Bernstein-Bezier function; on the basis, constructing a horizontal wave channel as a basic structure of the cold plate, optimizing the horizontal wave channel by adopting an NSGA-II optimization algorithm by taking the average temperature and the root mean square temperature of the upper surface and the lower surface of the cold plate as objective functions, carrying out normalization treatment on design variables, and finally selecting the best compromise solution by a TOPSIS sorting method. The invention relates to a free modeling and optimizing design method of a horizontal wave channel, which aims to eliminate the limitation of a fixed geometric structure form and a size on heat transfer performance, lays out the structural distribution of a horizontal wave channel cold plate and improves the heat dissipation performance of the horizontal wave channel cold plate.
Description
Technical Field
The invention belongs to the technical field of high-power electronic liquid cooling radiator methods, and particularly relates to a free modeling and optimal design method for a horizontal wavy channel.
Background
With the continuous development of the electronic industry, high-gain, high-density and miniaturized electronic devices have become a major trend of current development. Electronic thermal management has received a great deal of attention as an indispensable link. In recent years, in order to meet the heat dissipation demand, efficient and rapid cooling systems have been studied to improve heat transfer efficiency. At present, air cooling is widely used as a main cooling method. However, such a heat sink is not suitable for small-sized high-power electronic devices due to the low thermal conductivity and heat capacity of air, and thus a new approach is required. Liquid-cooled heat sinks have superior performance compared to air and will be used in different cold plate designs depending on the different strategies for fluid flow and heat transfer characteristics. Various runner forms and cross-sectional shapes exist in the overall design of the cold plate, wherein the runner forms comprise serpentine, parallel, tree-shaped, convergent or divergent runners, and the cross-sectional shapes comprise circular, rectangular, triangular and the like. Previous studies have shown that channels employing a laterally perturbed configuration can be considered vortex generators such as building cavities, adding wings (ribs or fins), using torsion bands, designing corrugated microchannels, and the like. These methods can induce the generation of lateral and longitudinal vortices, re-developing fluid flow, and thereby improving heat transfer coefficients. In particular, with respect to corrugated channels, recent researches have been increasing, which show superiority in heat transfer performance, mainly due to the presence of a backflow zone and secondary flow to remix fluid flow, re-develop a thermal boundary layer, and thus improve heat exchange capability. In addition, by increasing the corrugation, the corrugation channels can significantly increase the heat transfer area in a compact space, the geometrical features of which include amplitude, wavelength, waveform and cross-sectional shape and size, and the different arrangement of these geometrical features can lead to variations in the shape and size and strength of the recirculation zone. Therefore, in re-analyzing the corrugated channels, different conditions and applications need to be considered to determine the impact of the geometric distribution and dimensions of the corrugated channels on the cold plate design. The problems of the prior art are as follows: 1) Current horizontal undulating channel designs focus only on two-dimensional configurations, followed by simple stretching to obtain three-dimensional undulating configurations. 2) The geometric features of the corrugated channels are of a fixed form and size, which is detrimental to further exploration of their heat transfer laws and improvement of heat dissipation performance. 3) Since the local structural features involved in most designs are limited, this weakens the potential of the geometry for enhanced heat transfer performance, and the design results obtained are more conventional and heat exchange efficiency limited. 4) The design process depends on experience and inspiration of the designer, and whether the final result obtained by the whole design process is optimal is still further questionable. 5. The design of the wave-shaped channel involves the adjustment and optimization of a plurality of parameters, a great deal of simulation and experimental study is needed, and the design complexity is high. 6) Conventional fixed-shape-sized corrugated channel designs are suitable for certain specific fluid flow conditions and have poor applicability to other fluid flow conditions.
Disclosure of Invention
The invention aims to provide a free modeling and optimal design method for a horizontal wave channel, which aims to eliminate the limitation of the heat transfer performance of a fixed geometric structure form and a size, lay out the structural distribution of a horizontal wave channel cold plate and improve the heat dissipation performance of the horizontal wave channel cold plate.
The technical scheme adopted by the invention is as follows: the free modeling and optimizing design method of the horizontal wave channel adopts a cubic spline function to represent amplitude, a wavelet function to represent wavelength and a truncated function to represent waveform, and realizes the free correction of the shape and the size of the cross section of the channel through a rational Bernstein-Bezier function; on the basis, constructing a horizontal wave channel as a basic structure of the cold plate, optimizing the horizontal wave channel by adopting an NSGA-II optimization algorithm by taking the average temperature and the root mean square temperature of the upper surface and the lower surface of the cold plate as objective functions, carrying out normalization treatment on design variables, and finally selecting the best compromise solution by a TOPSIS sorting method.
The invention is also characterized by comprising the following steps:
step 1, determining the external dimension, hydraulic diameter, heating surface dimension and runner inlet and outlet positions of a horizontal wavy channel cold plate according to the configuration condition of field electronic equipment;
step 2, determining geometric parameters of an inlet and an outlet of a flow channel according to the determined outline dimension of the cold plate, determining heat transfer attribute parameters according to the dimension of a heating surface, determining flow attribute parameters according to the hydraulic diameter and the positions of the inlet and the outlet of the flow channel, and determining material attributes of a heat dissipation system according to engineering requirements;
step 3, constructing an initial straight channel with any cross-sectional shape based on a rational Bernstein-Bezier function through rotation and translation matrix parameterization according to the determined outline dimension of the cold plate;
step 4, constructing geometrical characteristics of amplitude, wavelength and waveform of the horizontal wave channel according to the initial straight channel and based on any shape parameterized modeling method of the wave channel;
step 5, establishing a cold plate finite element model based on a conjugate heat transfer control equation according to the horizontal wave-shaped channel;
step 6, according to the cold plate finite element model, grid independence test is executed, and the reliability of the numerical model is verified;
step 7, normalizing four design variables of amplitude, wavelength, waveform, section shape and size corresponding to the geometric features;
step 8, constructing design variables of an optimization framework according to engineering requirements, wherein the design variables comprise control coefficients of cubic spline functions representing amplitudes, wavelet functions representing wavelengths, truncation functions representing waveforms and control coefficients of Bernstein functions representing channel sections and channel torsion; constructing an objective function of the optimization framework, including minimizing the average temperature and the root mean square temperature of the top and the bottom of the cold plate;
step 9, optimizing the objective function of the optimization framework obtained in the step 8 by adopting an NSGA-II optimization algorithm to obtain the optimized temperature distribution of the cold plate and the power device;
and 10, selecting the optimal compromise solution by a TOPSIS sorting method according to the numerical solution of the temperature distribution of the optimized cold plate and the power device.
In the step 2, the external dimensions of the cold plate comprise length L, width W and height H; hydraulic diameter D; the heating surface has a size W h ×L h The method comprises the steps of carrying out a first treatment on the surface of the The heat transfer attribute parameter includes an inlet fluid temperature T in An insulated boundary of an external wall, discrete heat sources Q; the flow attribute parameter includes an inlet flow rate U in And outlet pressure P out The method comprises the steps of carrying out a first treatment on the surface of the The heat dissipation system is made of solid materials aluminum 6061-T6 and liquid transportation working medium water, and the properties of the heat dissipation system include: solid thermal conductivity k s Thermal conductivity of fluid k f Constant pressure specific heat capacity C of fluid p And a fluid density ρ.
In step 3, an initial straight-channel architecture is constructed by rotation and translation matrix parameterization, fitting is performed in the XOY plane, and Z is the height direction thereof.
In step 4, the basic architecture of the horizontal wave-like channel model is primarily determined by studying four design variables, namely amplitude, wavelength, waveform, cross-sectional shape and size.
In step 5, a control equation of conjugate heat transfer is established, and boundary conditions are defined.
In step 6, the grid independence is checked, and the finite element model is subjected to grid division by adopting unstructured tetrahedral grids.
In step 7, the design variables of the horizontal wave-shaped channel are normalized, so that the model architecture of the horizontal wave-shaped channel is further obtained.
In step 8, an optimization framework is built regarding the design variables, objective functions, and constraints.
In step 9, a non-dominant ranking genetic algorithm II (NSGA-II) is adopted, a corresponding population number P and an iteration number G are set, and a related convergence criterion is set.
In step 10, the optimized model is further discussed by selecting the best compromise solution by the approach of similarity to ideal solution rank order preference Technique (TOPSIS).
The beneficial effects of the invention are as follows:
1) The invention can effectively perform free modeling on the horizontal corrugated structure model, simultaneously considers four design variables of amplitude, wavelength, waveform, section shape and size, overcomes the limitation of fixed geometric size and shape, and realizes the design of free molded surfaces.
2) The invention allows for the amplitude to be characterized by a cubic spline function, primarily because the cubic spline function can pass through the points and has a continuous first derivative, which is beneficial to solving the problem of sensitivity to a given point (e.g., amplitude boundary). Meanwhile, the waveform is represented by adopting the modified wavelet function, so that the continuity of the wavy channel can be ensured. Since the Bernstein function can handle the global trend of the generated Bezier curve with only a small number of control points, and redundant design variables are not required, which is advantageous in constructing an efficient and compact optimization system, the construction of cross-sectional shapes and sizes using the Bernstein function is considered.
3) The invention defines each design variable by adopting a corresponding variable function, further improves the flexibility of the whole modeling process, overcomes the limitation of the traditional modeling on the size and shape, and has more universality.
4) The invention provides a horizontal corrugated channel design scheme based on a parameterized modeling system, and the heat dissipation performance of the horizontal corrugated channel is greatly improved by applying an optimization technology, so that the reliability of the horizontal corrugated channel is ensured. The scheme is not only suitable for the cold plate radiator, but also can be popularized to equipment such as the radiator with the fins, the heat exchanger and the like, and has certain guiding significance on the design of the high-power electronic liquid cooling radiator.
Drawings
FIG. 1 is a flow chart of the horizontal wave channel free modeling and optimization design method of the present invention;
FIG. 2 is an initial straight channel 3D model;
FIG. 3 is a 3D model of a horizontal undulating channel for comparison and experience-based;
FIG. 4 is a Pareto front plot with minimum surface average temperature versus root mean square temperature as an objective function;
FIG. 5 is a temperature distribution of upper and lower surfaces of a straight channel, a uniform wave channel, a variable wave channel;
FIG. 6 is a temperature profile of the middle layer in the height direction of a straight channel, a uniform wavy channel, and a variable wavy channel;
fig. 7 is a streamline profile of a straight channel, a uniform wavy channel, a variable wavy channel.
Detailed Description
The invention will be described in detail with reference to the accompanying drawings and detailed description.
The invention provides a free modeling and optimizing design method of a horizontal wave channel, which aims to eliminate the limitation of a fixed geometric structure form and a size on heat transfer performance, lays out the structural distribution of a horizontal wave channel cold plate through an optimizing technology, and further improves the heat dissipation performance of the horizontal wave channel cold plate by taking the minimum average temperature and the minimum root mean square temperature of an upper plane and a lower plane as optimizing objective functions. The invention further explores the potential of the horizontal corrugated structure in improving the heat radiation performance of the radiator, and lays a theoretical foundation for related engineering practice.
Example 1
The horizontal wave channel free modeling and optimizing design method provided by the invention is based on parameterized modeling, and allows the wave channel to be freely constructed in the aspects of amplitude, wavelength, waveform and section, so that fixed forms and sizes are liberated. Specifically, the invention adopts a cubic spline function to represent the amplitude, a wavelet function to represent the wavelength and a newly defined truncated function to represent the waveform, and realizes the free correction of the cross section shape and the size of the channel through a Bernstein function. On the basis, the invention constructs the horizontal corrugated channel which can be used as the basic structure of the cold plate. Aiming at the optimization problem of the horizontal corrugated channel, the average temperature and the root mean square temperature of the upper surface and the lower surface of the cold plate are used as target functions, and a non-dominant order genetic algorithm (NSGA-II) is adopted for optimization, and meanwhile, normalization processing is carried out on design variables so as to improve the stability of optimization. Finally, for the optimized model, selecting the best compromise solution by a similarity ordering preference Technique (TOPSIS) method with an ideal scheme. The work can be applied not only to cold plates, but also to parts with wave structures such as finned radiators, heat exchangers and the like.
Example 2
As shown in FIG. 1, the method for freely modeling and optimizing the design of the horizontal wave-shaped channel provided by the invention comprises the following specific steps:
step 1, determining the surface heat flux density of a power device, the basic outline dimension of a cold plate and the position of an inlet and an outlet of a runner
Determining the surface heat flux density Q of the power device according to the configuration condition of the field electronic equipment; parameters of the external dimensions of the cold plate: length L, width W, height H; the location of the doorway is shown in fig. 2.
Step 2, determining physical boundary parameters and corresponding physical parameters of the cold plate
Determining inlet geometric parameters including inlet diameter D according to the determined heat flux density and the cold plate external dimension in and Dout The method comprises the steps of carrying out a first treatment on the surface of the The heat transfer attribute parameter includes an inlet fluid temperature T in An insulated boundary of an external wall, discrete heat sources Q; the flow attribute parameter includes an inlet flow rate U in And outlet pressure P out The method comprises the steps of carrying out a first treatment on the surface of the The heat dissipation system is made of solid materials aluminum 6061-T6 and liquid transportation working medium water, and the properties of the heat dissipation system include: solid thermal conductivity k s Thermal conductivity of fluid k f Constant pressure specific heat capacity C of fluid p And a fluid density ρ.
Step 3, determining an initial straight channel
An initial straight channel is constructed through rotation and translation matrix parameterization, fitting is performed in an XOY plane, and Z is the height direction. The rotation and translation matrix is:
wherein X (t, θ), Y (t, θ), Z (t, θ) in item 1 are spatial locations of the corresponding geometric feature in the full coordinate system; items 2 and 3 represent geometric feature space positions x (t, θ), Y (t, θ), Z (t, θ) established under the local coordinate system (item 4), item 2 being rotation about the Z axis (determining that the cross section must be perpendicular to the channel centerline), and item 3 being rotation about the Y axis (determining the channel torsion degree); item 4 relates to cross-sectional shape and size, the meaning of the control parameters being different in different cross-sectional types; item 5 is the amount of local to global translation. Wherein any section type is constructed from a rational Bernstein-Bezier function, i.e
wherein ,Bi (t, θ) is built on a three-point basis; basis function b j,2 (θ) and weighting coefficientIn combination, the shape of the cross-section is determined, in order to create a different cross-sectional shape along the length of the channel,/->Can be expressed as Bernstein-Bezier functions
wherein ,is Bernstein-Bezier basis function; />Is a binomial coefficient; />The control coefficient of the intermediate parameter of the ith arc line; n is n w Is the maximum power of the function; the process is thatThe free construction of the cross section shape can be realized by only using the design variables of the intermediate parameters, so that a large number of design variables in the traditional design are avoided, and the subsequent establishment of a compact optimization system is facilitated.
Step 4, constructing a horizontal corrugated channel on the basis of the initial straight channel
Reconverting x (t, θ), z (t, θ) on the basis of the initial straight channel to obtain
(1) wf (t) is a trigonometric function of the parameter t, which yields a wavy path expressed as
wf(t)＝amp(t)g(t)＝amp(t)cos(freq(t)πt)
Wherein amp (t) is the amplitude of the wavy channel; freq (t) is the wavelength of the wavy channel.
(2) The amp (t) can be built by a cubic spline function, which is not an essential component of the modeling of the wavy channel, but is advantageous in solving the problem of sensitivity to a given point (e.g., amplitude boundary) because the function can pass through these points and has a continuous first derivative. The simplified cubic spline function expressing the amplitude is
amp(t)＝f(amp 1 ,…,amp j ,…,amp na ,t),(0≤t≤1)
wherein ,ampj Is a control coefficient of the amplitude; n is n a Is the maximum value.
(3) In the expression of wf (t), because g (t) is more complex and the accuracy requirement on geometric modeling is possibly higher, a Haar wavelet function needs to be introduced to reconstruct g (t), and the wavelet function and the expression of g (t) are respectively
(4) Since both H (t) endpoints are in interval t E [0,1/n ] h ]The values of the above are all 1, when the value is accumulated to 2, the continuity of the wavy channel is not favored, and therefore, the original Haar wavelet function needs to be modified, namely
Wherein RH (t) is a modified Haar wavelet function; the general quadratic function pi (t) can be obtained by a given boundary.
(5) Taking the difference of waveforms into consideration, a new waveform function is introduced, the expression of which is
wherein ,Sj (t)＝cos(freq j Pi t); lambda (t) is a truncated function and different waveforms can be obtained by adjusting lambda (t).
(6) In order to ensure rapid modeling, the invention sets the minimum value of lambda (t) to 0.2, and lambda (t) is still defined by a cubic spline function, which can be expressed as
λ(t)＝f(λ 1 ,…,λ j ,…,λ nλ ,t),(0≤t≤1)
wherein ,λj Is the control coefficient of the waveform; n is n λ Is the maximum number of waveforms.
Step 5, determining a control equation according to the shape optimization criterion and giving corresponding boundary conditions
To simplify the analysis steps, the conjugate heat transfer control equation is determined as follows:
continuity equation:
momentum equation:
energy equation in the solid domain:
energy equation in the fluid domain:
wherein ,u1 ,u 2 ,u 3 Fluid velocity in the direction X, Y, Z, respectively; t is a temperature field; p is the pressure field; ρ, μ, C p The density, the dynamic viscosity and the specific heat capacity of the fluid are respectively; k (k) s ，k f Thermal conductivity of solids and fluids, respectively.
Step 6, normalizing the design variables
And according to the determined geometric characteristics, carrying out normalization processing on the four design variables of the corresponding amplitude, wavelength, waveform, section shape and size, and finally establishing a horizontal wavy channel model with the waveform and channel section capable of being changed at will.
Step 7, establishing a shape optimization criterion
The optimization framework is established as follows:
Find:ω
Subject to:0≤ω≤1
wherein ω is a normalized vector set of all the corresponding design variables, including a control coefficient characterizing the magnitude of the cubic spline function, a wavelet function characterizing the wavelength, a newly defined truncated function characterizing the waveform, and a bernstein function characterizing the channel cross section and channel twist; t is the average temperature of the upper surface, RMST is the root mean square temperature of the lower surface, Ω is the integral field, T is the temperature field of the field, dΩ represents the differential from the designated surface, and ε is the integral symbol.
Step 8, obtaining the optimal horizontal corrugated structure distribution and outputting a result
And optimizing an objective function by adopting a non-dominant order genetic algorithm (NSGA-II) according to the finite element model of the cold plate, and obtaining the improved temperature distribution of the designated surface (the surface on which the power device is placed) of the cold plate.
Optimizing objective function by non-dominant order genetic algorithm (NSGA-II), setting corresponding population number P and iteration number G, and setting related convergence criterion as follows
Wherein p is population; gen is the current iteration number; this means that the iteration will end when the target value of the last six iterations does not change.
Step 9, further discussing the optimized model by adopting a' approximate ideal solution ordering method (TOPSIS)
The best compromise solution is selected by a similarity to ideal scheme ordering preference Technique (TOPSIS) method, and the flow is as follows:
(1) Creating a matrix with m=30, n=20 (T ij ) m×n As an objective function, where T ij The average temperature and the root mean square of the temperature of the upper surface and the lower surface of the cold plate after the optimization solution are carried out;
(2) Normalizing the initial matrix
(3) By introducing weighting factors w j Weighting normalized matrices, i.e.
a ij ＝w j ×t ij
(4) Definition of positive and negative ideal solutions as
A + ＝(min[a 11 ,…,a m1 ],min[a 12 ,…,a m2 ],…,min[a 1n ,…,a nm ])
A - ＝(max[a 11 ,…,a m1 ],max[a 12 ,…,a m2 ],…,max[a 1n ,…,a nm ])
(5) Calculating the distance between the alternative and the positive/negative ideal solution as
(6) Definition of relative fitness c i Is that
(7) Selecting the best compromise
A best ＝A∈max(c i )
Example 3 (simulation case):
1. simulation parameters
Referring to fig. 2, the three-dimensional external dimension of the cold plate is 20mm 80mm 10mm, the hydraulic diameter is 6mm, the heated surface dimension is 10mm 60mm, and the heat flow density q at the inlet is equal to or higher than that of the cold plate in 16.67W/cm 2 Inlet temperature T in 293.15K, pressure p at outlet 0, inlet speed u 1 ＝u 3 ＝0,The reynolds number is 1000. The solid material is aluminum 6061-T6, and the liquid working medium is water.
2. Simulation content and results
FIG. 5 shows a straight channel cold plate (SC), a uniform wave coldPlate (UHWC), variable wave cold plate (VHWC-22) upper and lower surface temperature profiles. Wherein UHWC as shown in FIG. 3 is an empirically based comparative case, and VHWC-22 is obtained based on the acquisition of the pareto front graph (FIG. 4) and TOPSIS methods. Obviously, the temperature distribution of the upper and lower surfaces of each design is almost symmetrical, which means that the cold plate can be placed freely without regard to layout. As can be seen from the temperature distribution of the target surface in fig. 5, the convective heat transfer capability gradually becomes worse along the channel length, and thus the temperature also gradually increases along the channel length, resulting in a high temperature being mainly concentrated in the latter half. In combination with the temperature distribution of the intermediate layer in the SC height direction shown in fig. 6, a large temperature gradient between the solid and fluid domains can be found, which means that the heat does not flow well out of the cold plate system with the fluid. The temperature profile of UHWC is better than SC, and the high temperature area in the latter half becomes smaller. Maximum temperature of UHWC (T max ) Reduced by 5K, maximum temperature (T max ) And the lowest temperature (T) min ) The difference (Δt) of (c) is reduced by 4K. Although the average knoop-to-ser number (42.15) of UHWC is less than the average knoop-to-ser number (43.31) of the straight channel, the periodic enhancement of convection capability results in better overall thermal performance than SC with good temperature uniformity. The temperature profile of the intermediate layer in the thickness direction also shows that the fluid domain, cold fluid region, of UHWC is smaller than SC and more heat is given to the fluid. In addition, the temperature distribution of the optimized VHWC-22 target surface is obviously improved, and most areas show cooler contours, so that the variable horizontal corrugated channel is more beneficial to improving the heat dissipation performance. Maximum temperature (T) max ) The temperature difference (DeltaT) is reduced by at most 41K and at most 30K. VHWC-22 changes the original circular cross section into a smaller diamond-like cross section with a straight channel-like arrangement resulting in a reduction in convective heat transfer area. The average noose number of VHWC-22 was increased by 28.55 (65.92%) over SC and 29.71 (70.49%) over UHWC. Based on the temperature distribution of the intermediate layer of VHWC-22 in the thickness direction (as shown in FIG. 6), it was found that the temperature difference between the solid domain and the fluid domain becomes smaller, which further illustrates the superiority of the variable horizontal wave channel. The straight channel, uniform wave channel, variable wave channel streamline profile shown in FIG. 7 shows that UHWC and VHWC-22 channels produced compared to SCA large amount of rotating low velocity fluid. In addition, UHWC has the same maximum speed as SC (V max ) The value of 0.33m/s, although the cross-sectional area of the former is larger than that of the latter. VHWC-22 has a maximum speed value of 0.62m/s, mainly due to its smaller cross section. Combining the wave properties, this makes more high velocity fluid visible at the solid-liquid interface, meaning that the wave-like channels created facilitate the breaking of the flow boundary layer. Fig. 4 shows Pareto front distribution of average temperature and root mean square temperature of the objective function, and correlation values of the last 30 individuals have been normalized for more visual display, and the optimal structure is obtained by setting the ratio of average temperature to root mean square weight of temperature to 0.5:0.5.
Table 1 comparison of heat sink surface temperatures before and after optimization
The optimized variable corrugated channel has obviously better heat dissipation performance than the straight channel and the uniform corrugated channel through comparing and analyzing the surface temperature distribution of the three channels, and the heat dissipation performance is greatly improved.
Claims (10)
1. The free modeling and optimizing design method of the horizontal wave channel is characterized in that a cubic spline function is adopted to represent amplitude, a wavelet function is adopted to represent wavelength, a truncated function is adopted to represent waveform, and free correction of the cross section shape and the dimension of the channel is realized through a rational Bernstein-Bezier function; on the basis, constructing a horizontal wave channel as a basic structure of the cold plate, optimizing the horizontal wave channel by adopting an NSGA-II optimization algorithm by taking the average temperature and the root mean square temperature of the upper surface and the lower surface of the cold plate as objective functions, carrying out normalization treatment on design variables, and finally selecting the best compromise solution by a TOPSIS sorting method.
2. The method for freely modeling and optimizing design of horizontal wave-like channel as claimed in claim 1, comprising the steps of:
step 1, determining the external dimension, hydraulic diameter, heating surface dimension and runner inlet and outlet positions of a horizontal wavy channel cold plate according to the configuration condition of field electronic equipment;
step 2, determining geometric parameters of an inlet and an outlet of a flow channel according to the determined outline dimension of the cold plate, determining heat transfer attribute parameters according to the dimension of a heating surface, determining flow attribute parameters according to the hydraulic diameter and the positions of the inlet and the outlet of the flow channel, and determining material attributes of a heat dissipation system according to engineering requirements;
step 3, constructing an initial straight channel with any cross-sectional shape based on a rational Bernstein-Bezier function through rotation and translation matrix parameterization according to the determined outline dimension of the cold plate;
step 4, constructing geometrical characteristics of amplitude, wavelength and waveform of the horizontal wave channel according to the initial straight channel and based on any shape parameterized modeling method of the wave channel;
step 5, establishing a cold plate finite element model based on a conjugate heat transfer control equation according to the horizontal wave-shaped channel;
step 6, according to the cold plate finite element model, grid independence test is executed, and the reliability of the numerical model is verified;
step 7, normalizing four design variables of amplitude, wavelength, waveform, section shape and size corresponding to the geometric features;
step 8, constructing design variables of an optimization framework according to engineering requirements, wherein the design variables comprise control coefficients of cubic spline functions representing amplitudes, wavelet functions representing wavelengths, truncation functions representing waveforms and control coefficients of Bernstein functions representing channel sections and channel torsion; constructing an objective function of the optimization framework, including minimizing the average temperature and the root mean square temperature of the top and the bottom of the cold plate;
step 9, optimizing the objective function of the optimization framework obtained in the step 8 by adopting an NSGA-II optimization algorithm to obtain the optimized temperature distribution of the cold plate and the power device;
and 10, selecting the optimal compromise solution by a TOPSIS sorting method according to the numerical solution of the temperature distribution of the optimized cold plate and the power device.
3. The method for freely modeling and optimizing design of horizontal corrugated channel as claimed in claim 2, wherein in said step 2, the external dimensions of the cold plate include length L, width W, and height H; hydraulic diameter D; the heating surface has a size W h ×L h The method comprises the steps of carrying out a first treatment on the surface of the The heat transfer attribute parameter includes inlet fluid temperature T in An insulated boundary of an external wall, discrete heat sources Q; the flow attribute parameter includes an inlet flow velocity U in And outlet pressure P out The method comprises the steps of carrying out a first treatment on the surface of the The heat dissipation system material has solid material aluminum 6061-T6 and liquid transportation working medium water, and the heat dissipation system material attribute includes: solid thermal conductivity k s Thermal conductivity of fluid k f Constant pressure specific heat capacity C of fluid p And a fluid density ρ.
4. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein said step 3 specifically comprises: constructing an initial straight channel based on a rotation and translation matrix of a three-dimensional Cartesian coordinate system, fitting in an XOY plane, wherein Z is the height direction, and the rotation and translation matrix is as follows:
wherein X (t, θ), Y (t, θ), Z (t, θ) in item 1 are spatial locations of the corresponding geometric feature in the full coordinate system; items 2 and 3 represent geometric feature space positions x (t, θ), Y (t, θ), Z (t, θ) established under the local coordinate system, i.e., item 4, item 2 being rotated about the Z axis, and item 3 being rotated about the Y axis; item 4 relates to cross-sectional shape and size, the meaning of the control parameters being different in different cross-sectional types; item 5 is the local to global translation amount; wherein, any section type is constructed by a rational Bernstein-Bezier function, namely:
wherein ,Bi (t, θ) is built on a three-point basis; basis function b j,2 (θ) and weighting coefficientIn combination determine the shape of the cross section, +.>Expressed as Bernstein-Bezier functions:
wherein ,is Bernstein-Bezier basis function; />Is a binomial coefficient; />The control coefficient of the intermediate parameter of the ith arc line; n is n w Is the maximum power of the function.
5. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein said step 4 specifically comprises: horizontal undulating channels on the basis of the initial straight channel construction:
wherein , and />X (t, theta) and z (t, theta) are reconverted on the basis of the initial straight channel; wf (t) is a trigonometric function with respect to the parameter t, expressed as:
wf(t)＝amp(t)g(t)＝amp(t)cos(freq(t)πt)
wherein freq (t) is the wavelength of the wavy channel; amp (t) is the amplitude of the wavy channel, established by a cubic spline function, which expresses the simplified cubic spline function of amplitude as:
wherein ,ampj Is a control coefficient of the amplitude; n is n a Is at a maximum value;
in the expression of wf (t), a Haar wavelet function is introduced to reconstruct g (t), and the expressions of the wavelet function and g (t) are respectively:
when two endpoints of H (t) are in interval t E [0,1/n ] h ]When the above value is accumulated to 2, the original Haar wavelet function is modified, namely:
wherein RH (t) is a modified Haar wavelet function; the quadratic function pi (t) is obtained by a given boundary;
regarding the geometric features of the waveform, a new waveform function is introduced, expressed as:
wherein ,Sj (t)＝cos(freq j Pi t); lambda (t) is a truncated function, and different waveforms are obtained by adjusting lambda (t); let λ (t) be 0.2, defined by a cubic spline function, expressed in simplified form as:
λ(t)＝f(λ 1 ,…,λ j ,…,λ nλ ,t),(0≤t≤1)
wherein ,λj Is the control coefficient of the waveform; n is n λ Is the maximum number of waveforms.
6. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein in step 5, the conjugated heat transfer control equation is determined as follows:
continuity equation:
momentum equation:
energy equation in the solid domain:
energy equation in the fluid domain:
wherein ,u1 ,u 2 ,u 3 Fluid velocity in the direction X, Y, Z, respectively; t is a temperature field; p is the pressure field; ρ, μ, C p The density, the dynamic viscosity and the specific heat capacity of the fluid are respectively; k (k) s ，k f The thermal conductivity coefficients of the solid and the fluid respectively;
on the basis, a cold plate finite element model with horizontal wave-shaped channels is built.
7. The method for freely modeling and optimizing design of horizontal wave channel according to claim 2, wherein in the step 6, when grid independence test is performed, the grids of the fluid domain are encrypted by unstructured grids, the obtained number of grids is compared, and finally an optimized grid is selected.
8. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein the optimization framework established in the step 8 is as follows:
Find:ω
Minimize:
Subjectto:0≤ω≤1
wherein ω is a normalized vector set of all the corresponding design variables, including a control coefficient characterizing the magnitude of the cubic spline function, a wavelet function characterizing the wavelength, a newly defined truncated function characterizing the waveform, and a bernstein function characterizing the channel cross section and channel twist;RMST is the root mean square temperature of the lower surface, Ω is the integral field, T is the temperature field of the field, dΩ represents the differential from the designated surface, and Σ is the integral symbol.
9. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein in step 9, NSGA-II optimization algorithm is adopted to optimize objective function, corresponding population number P and iteration number G are set, and related convergence criteria are set as follows:
wherein p is population; gen is the current iteration number; when the target values of the last six iterations are unchanged, the iterations are terminated; and then, according to a convergence criterion, performing optimization to obtain the temperature distribution of the superior horizontal wavy channel and the power device.
10. The method for freely modeling and optimizing design of horizontal wave channel as claimed in claim 2, wherein said step 10 comprises the steps of:
step 10.1, creating a matrix with m=30, n=20 (T ij ) m×n As an objective function, where T ij The average temperature and the root mean square temperature of the top and the bottom of the optimized cold plate are obtained;
step 10.2, carrying out normalization processing on the initial matrix;
step 10.3 by introducing a weighting factor w j Weighting the normalized matrix, namely:
a ij ＝w j ×t ij
step 10.4, defining positive and negative ideal solutions as:
A + ＝(min[a 11 ,…,a m1 ],min[a 12 ,…,a m2 ],…,min[a 1n ,…,a nm ])
A - ＝(max[a 11 ,…,a m1 ],max[a 12 ,…,a m2 ],…,max[a 1n ,…,a nm ])
step 10.5, calculating the distance between the alternative solution and the positive/negative ideal solution as follows:
step 10.6, defining the relative fitness c i The method comprises the following steps:
step 10.7, selecting the best compromise:
A best ＝A∈max(c i )。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310624959.1A CN116796593A (en) | 2023-05-30 | 2023-05-30 | Free modeling and optimal design method for horizontal wave-shaped channel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310624959.1A CN116796593A (en) | 2023-05-30 | 2023-05-30 | Free modeling and optimal design method for horizontal wave-shaped channel |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116796593A true CN116796593A (en) | 2023-09-22 |
Family
ID=88038982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310624959.1A Pending CN116796593A (en) | 2023-05-30 | 2023-05-30 | Free modeling and optimal design method for horizontal wave-shaped channel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116796593A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117117248A (en) * | 2023-10-23 | 2023-11-24 | 佛山市清极能源科技有限公司 | Method for optimizing bipolar plate flow channel structure |
-
2023
- 2023-05-30 CN CN202310624959.1A patent/CN116796593A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117117248A (en) * | 2023-10-23 | 2023-11-24 | 佛山市清极能源科技有限公司 | Method for optimizing bipolar plate flow channel structure |
| CN117117248B (en) * | 2023-10-23 | 2024-01-23 | 佛山市清极能源科技有限公司 | Method for optimizing bipolar plate flow channel structure |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Freegah et al. | CFD analysis of heat transfer enhancement in plate-fin heat sinks with fillet profile: Investigation of new designs | |
| Menni et al. | Combination of baffling technique and high-thermal conductivity fluids to enhance the overall performances of solar channels | |
| CN111709096B (en) | Design method of special-shaped fin structure for strengthening natural convection heat transfer | |
| CN109800507A (en) | A kind of pair of secondary Structural shape optimization of heat dissipation cold plate topology boundary | |
| Pourdel et al. | Numerical investigation of turbulent flow and heat transfer in flat tube: effect of dimples with operational goals | |
| Yang et al. | Forced convective cooling of a fin in a channel | |
| CN116796593A (en) | Free modeling and optimal design method for horizontal wave-shaped channel | |
| CN108256272B (en) | S-shaped flow passage layout optimization design method for liquid cooling radiator | |
| Xie et al. | Topology optimization of heat sink based on variable density method | |
| Yan et al. | Numerical investigation on thermal-hydraulic characteristics of the micro heat sink with gradient distribution pin fin arrays and narrow slots | |
| Ameur et al. | Effect of some parameters on the thermohydraulic characteristics of a channel heat exchanger with corrugated walls | |
| Feng et al. | Constructal design for tree-shaped compound heat transfer channel in a disc heat generation body | |
| Soleimani et al. | Multi-objective optimization of 3D micro-fins using NSGA-II | |
| Yeranee et al. | Enhanced thermal performance of a pin-fin cooling channel for gas turbine blade by density-based topology optimization | |
| Hu et al. | Optimization of bionic heat sinks with self-organized structures inspired by termite nest morphologies | |
| Sun et al. | An intelligent plate fin-and-tube heat exchanger design system through integration of CFD, NSGA-II, ANN and TOPSIS | |
| Meng et al. | Performance enhancement for exhaust thermoelectric power generation system by using porous pin fins based on a fully automatic optimization method | |
| Yu et al. | Efficient optimization of parallel micro-channel heat sinks based on flow resistance network model | |
| Li et al. | On identifying optimal heat conduction topologies from heat transfer paths analysis | |
| Erdinc et al. | An optimization study on thermo-hydraulic performance arrays of circular and diamond shaped cross-sections in periodic flow | |
| Liu et al. | Influence of fin arrangement on fluid flow and heat transfer in the inlet of a plate-fin heat exchanger | |
| CN114757125A (en) | Self-organizing heat sink structure design method based on diffusion back-diffusion system | |
| Ahmadian-Elmi et al. | Investigation of the optimal configuration of a highly conductive material embedded in a triangular fin | |
| Jiang et al. | Optimization of micro-channel heat sink with trapezoidal cavities and solid/slotted oval pins based on genetic algorithm and Back propagation neural network | |
| Qi et al. | Assessment of two different optimization principles applied in heat conduction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |