Multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD

Abstract

The invention discloses a liquid cooling radiator multi-objective optimization method based on response surface analysis and CFD. The method comprises the following steps: determining an optimization objective and an evaluation index according to the structural characteristics of a liquid cooling radiator; performing single factor analysis on the plurality of determined optimization targets by using a CFD technology in combination with the evaluation indexes so as to determine the optimization targets or the value range of the optimization targets; generating a plurality of response surface analysis test schemes and establishing corresponding radiator response surface models; performing numerical simulation on each established radiator response surface model, and calculating to obtain an evaluation index corresponding to each radiator response surface model; verifying a linear regression result of the radiator response surface model; and determining an optimal radiator response surface model based on response surface analysis and a linear regression result, and taking an optimization target or a value range of the optimization target corresponding to the radiator response surface model as an optimal design parameter combination.

Description

Technical Field

[0001] The invention belongs to the technical field of radiator optimization design, and in particular to a multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD. Background Art

[0002] In recent years, liquid cooling is the most mainstream battery thermal management mode. It uses coolant to absorb and take away the heat generated by the battery, and can also be applied to other high-heat-generating power devices. Compared with air cooling, phase change and other cooling methods, it has excellent heat dissipation performance and is widely used in mass-produced automobiles. In particular, fin-type liquid-cooled radiators have attracted more and more attention from scholars due to their low pressure loss and good heat dissipation performance. At present, the optimization design of liquid-cooled radiators mainly focuses on the optimization of their structural parameters on the maximum temperature and temperature difference of the battery; the influence of these structural parameters on the average temperature of the radiator is rarely considered. At the same time, the loss of flow resistance generated by the coolant during the flow process, that is, the power consumption of the pump, is also an important indicator affecting the thermal management performance of the battery, and has not been studied equally.

[0003] Therefore, a more comprehensive analysis method for design parameters is needed to obtain the optimal liquid cooling radiator design, shorten the R&D cycle and improve efficiency. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides a multi-objective optimization method for liquid cooling radiators based on response surface analysis and CFD. In the design process, the influence of radiator structural parameters on average temperature and pressure drop is comprehensively analyzed, and the comprehensive performance of heat transfer and resistance is evaluated through comprehensive performance indicators. By combining the response surface analysis method and CFD technology, the R&D cycle is shortened, invalid experimental design schemes are reduced, computer computing power is fully utilized, and the calculation time of invalid schemes is reduced, so as to obtain the best radiator optimization design parameters.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD comprises the following steps:

[0007] S1. Establish a physical model of the liquid cooling radiator, determine the optimization target and evaluation index according to the structural characteristics of the liquid cooling radiator; take the structural parameters of the radiator as the optimization target; take the average temperature, pressure drop, and comprehensive evaluation index of the radiator as the evaluation index;

[0008] S2. Using CFD technology combined with evaluation indicators, single factor analysis is performed on the determined multiple optimization targets to determine the optimization target or the value range of the optimization target;

[0009] S3, generating several response surface analysis test plans based on the optimization target or the value range of the optimization target determined in S2;

[0010] S4, based on several response surface analysis test schemes generated in S3, respectively establish the radiator response surface models corresponding to the schemes;

[0011] S5. numerically simulate each established radiator response surface model, and calculate the evaluation index corresponding to each radiator response surface model;

[0012] S6. Verify the linear regression results of the radiator response surface model;

[0013] S7. Determine the optimal radiator response surface model based on the response surface analysis and linear regression results, and set the optimization target or the value range of the optimization target corresponding to the radiator response surface model as the optimal design parameter combination.

[0014] Furthermore, the structural parameters of the radiator include the inlet and outlet design, size, and internal structure design of the radiator.

[0015] Furthermore, if the radiator is a fin-type liquid-cooled radiator, its internal structure is designed with fin arrangement, fin spacing ratio f / a and w / b, inlet and outlet arrangement, and outlet spacing L; the front and rear of the fins are divided inside the radiator according to the flow direction of the coolant, then f is the spacing between adjacent fins, w is the spacing between adjacent fins, a is the span length of the fin, and b is the chord length;

[0016] Furthermore, the fin arrangement methods include equal spacing arrangement, variable spacing arrangement, array alignment arrangement, and staggered arrangement.

[0017] Furthermore, the import and export arrangements include single inlet and single outlet, single inlet and multiple outlets, multiple inlets and single outlets, and multiple inlets and multiple outlets.

[0018] Furthermore, the method of single factor analysis in S2 is: for the determined multiple optimization objectives, one of the optimization objectives is selected as a variable, and the other optimization objectives are fixed; the evaluation indicators corresponding to different variables are calculated respectively, and the optimal variable is determined based on the evaluation indicators.

[0019] Further, the average temperature is expressed as:

[0020]

[0021] Where T is the surface temperature of the radiator, A W is the surface area of ​​the radiator, and A is the surface of the radiator.

[0022] Further, the pressure drop is expressed as:

[0023] ΔP=ΔP in -ΔPout

[0024] Among them, ΔP in is the inlet gauge pressure, ΔP out is the outlet gauge pressure.

[0025] Further, the comprehensive evaluation index is expressed as:

[0026] TP=Nu / f 1 / 3

[0027] Where Nu and f are the average Nusselt coefficient and the average friction coefficient, respectively.

[0028] Furthermore, several response surface analysis experimental schemes were obtained using Design-Experiment software.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] (1) Based on response surface analysis and CFD technology, the present invention takes into account the average temperature and pressure drop of the radiator, introduces a comprehensive evaluation performance index (TP) as an evaluation index of the comprehensive performance of the radiator, and reduces the average temperature and pressure drop of the radiator and improves the comprehensive performance of the radiator through multi-objective optimization of the fin radiator.

[0031] (2) The optimization design method of the present invention can effectively shorten the design cycle of the fin heat sink, reduce the development cost, and improve the comprehensive performance of the heat sink. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Flow chart of the method of the present invention.

[0033] Figure 2 Schematic diagram of various arrangements of the fins of a bionic shark skin tooth fin heat sink, where (a) is a conventional equidistant array arrangement, and (b) and (c) are equidistant staggered arrangements.

[0034] Figure 3 Schematic diagram of various inlet and outlet arrangements of the bionic shark skin tooth fin radiator, including (a) a single inlet and single outlet design, (b) a single inlet and double outlet design, and (c) a double inlet and double outlet design.

[0035] Figure 4 Schematic diagram of the fin arrangement of the bionic shark skin tooth fin radiator. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0037] A multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD comprises the following steps:

[0038] S1. Establish a physical model of the finned liquid cooling radiator, and determine the optimization objectives and evaluation indicators according to the structural characteristics of the finned liquid cooling radiator.

[0039] The structural parameters of the radiator are taken as the optimization target, and the structural parameters include the radiator inlet and outlet design, size, internal structure design, etc.

[0040] This embodiment takes a fin-type liquid-cooled radiator as an example. The fin-type liquid-cooled radiator is designed with an inlet and an outlet, and fins are arranged on its inner wall, and the fins are arranged on the inner wall according to a certain pattern. Therefore, the structural parameters of the fin-type liquid-cooled radiator specifically include the fin arrangement method, the fin spacing ratio f / a and w / b, the inlet and outlet arrangement method, and the outlet spacing L. It should be noted that the front and back of the fins are divided inside the radiator according to the flow direction of the coolant, then f is the spacing between adjacent fins on the left and right, w is the spacing between adjacent fins on the front and back, a is the span length of the fin, and b is the chord length. Fin arrangement methods include equal spacing arrangement, variable spacing arrangement, array alignment arrangement, staggered arrangement, etc. Inlet and outlet arrangements include single inlet and single outlet, single inlet and multiple outlets, multiple inlets and single outlets, multiple inlets and multiple outlets, etc.

[0041] Determine the evaluation indicators of the radiator, which include:

[0042] (1) The average temperature used to reflect the radiator temperature:

[0043]

[0044] Where T is the surface temperature of the radiator, A W is the surface area of ​​the radiator, and A is the surface of the radiator.

[0045] (2) Pressure drop to reflect the pump power required by the radiator:

[0046] ΔP=ΔP in -ΔP out

[0047] Among them, ΔP in is the inlet gauge pressure, ΔP out is the outlet gauge pressure;

[0048] (3) Comprehensive evaluation index used to reflect the comprehensive performance of the radiator:

[0049] TP=Nu / f 1 / 3

[0050] Where Nu and f are the average Nusselt coefficient and average friction coefficient, respectively;

[0051] S2. Use CFD technology combined with evaluation indicators to perform single factor analysis on the multiple optimization targets to determine the optimization target or the value range of the optimization target. The specific process is as follows:

[0052] S2.1, fixed fin structural parameters, f / a = 1, w / b = 1.4, selected single inlet and single outlet layout, and performed single factor analysis on different fin arrangements. The specific arrangement is as follows Figure 2 As shown in the figure, the three evaluation indicators under different arrangement modes are calculated respectively, and the arrangement mode is determined based on the evaluation indicators.

[0053] S2.2, fixed fin structural parameters, fin arrangement (such as staggered mode 1), f / a = 1, w / b = 1.4, single factor analysis of various inlet and outlet arrangements, the inlet and outlet arrangements are as follows Figure 3 The three evaluation indicators under different inlet and outlet layouts are calculated respectively, and the inlet and outlet layout is determined based on the evaluation indicators.

[0054] S2.3, fixed fin structural parameters, fin arrangement (such as staggered mode 1), w / b = 1.4, inlet and outlet arrangement (such as single inlet and double outlet mode), single factor analysis is performed on the model in the range of f / a = 0.7-1.2, as shown in the schematic diagram Figure 4 The three evaluation indicators under different f / a value ranges are calculated respectively, and the f / a value range is determined based on the evaluation indicators.

[0055] S2.4, fixed fin structural parameters, fin arrangement (such as staggered mode 1), f / a = 1, inlet and outlet arrangement (such as single inlet and double outlet mode), single factor analysis is performed on the model in the range of w / b = 1-1.5, as shown in the schematic diagram Figure 4 As shown in the figure, the three evaluation indicators under different w / b value ranges are calculated respectively, and the w / b value range is determined based on the evaluation indicators.

[0056] S2.5, structural parameters of fixed fins, tooth fin arrangement (select staggered mode 1), f / a = 1, w / b = 1.4, inlet and outlet arrangement (such as selecting single inlet and double outlet mode), single factor analysis is performed on the model with outlet spacing L = 24-29cm, as shown in the schematic diagram Figure 3 The three evaluation indicators under different value ranges of the outlet spacing L are calculated respectively, and the value range of the outlet spacing L is determined based on the evaluation indicators.

[0057] In the above process of determining the optimization target and its scope, different optimization targets have different effects on the evaluation indicators. Therefore, it is necessary to comprehensively consider the evaluation indicators of each optimization target, as follows:

[0058] When the bionic shark skin tooth fins are arranged in staggered arrangement 1, the radiator has a lower average temperature, so staggered arrangement 1 is selected for response surface analysis;

[0059] When the inlet and outlet layout is single-inlet and double-outlet, the radiator has a lower pressure drop and a lower average temperature, so the single-inlet and double-outlet layout is selected for response surface analysis;

[0060] When f / a=0.9, the heat sink has better comprehensive performance, so f / a is set to 0.8-1 for response surface analysis;

[0061] When w / b=1.4, the heat sink has better comprehensive performance, so w / b is 1.3-1.5 for response surface analysis;

[0062] When the outlet spacing L = 25 cm, the radiator has a lower pressure drop and better overall performance, so the outlet spacing L is 24-26 cm for response surface analysis.

[0063] S3, inputting the optimization target or the value range of the optimization target determined in S2 into the Design-Experiment software to obtain several response surface analysis test schemes, and the response surface analysis test table is shown in Table 1;

[0064] Table 1 Response surface analysis test table

[0065]

[0066]

[0067] S4, based on several response surface analysis test schemes generated in S3, respectively establish the radiator response surface models corresponding to the schemes;

[0068] S5. Numerical simulations are performed on each of the established radiator response surface models, and the evaluation indexes corresponding to each radiator response surface model are calculated. The numerical simulation results are shown in Table 2.

[0069] Table 2 Numerical simulation results

[0070] Serial number f / a w / b L(cm) T_ave(℃) ΔP TP 1 0.9 1.4 25 34.835 5.509 5.027 2 0.9 1.3 24 34.873 5.303 5.033 3 0.9 1.5 26 34.836 6.164 5.024 4 0.9 1.5 24 34.837 6.160 5.024 5 0.8 1.3 25 34.840 5.325 5.052 6 1 1.5 25 34.867 6.142 5.007 7 0.9 1.4 25 34.835 5.509 5.027 8 0.8 1.4 26 34.828 5.505 5.017 9 0.9 1.4 25 34.835 5.509 5.027 10 0.9 1.4 25 34.835 5.509 5.027 11 1 1.4 26 34.864 5.488 5.010 12 1 1.4 24 34.867 5.500 5.003 13 1 1.3 25 34.967 5.409 4.935 14 0.9 1.3 26 34.871 5.309 5.020 15 0.8 1.5 25 34.803 6.318 5.029 16 0.8 1.4 24 34.810 5.536 5.039 17 0.9 1.4 25 34.835 5.509 5.027

[0071] S6. According to the numerical simulation results shown in Table 2, the linear regression results of the radiator response surface model are verified. Table 3 shows the linear regression results of the response surface model of T_ave, ΔP and TP. Generally speaking, the closer the determination coefficient R2 is to 1, the more accurate the prediction result is. According to Table 3, the R2 of T_ave, ΔP and TP are 0.9669, 0.9997 and 0.9713 respectively, which verifies the reliability of the response surface model.

[0072] Table 3 Linear regression table

[0073]

[0074] S7. Determine the optimal radiator response surface model based on the response surface analysis and linear regression results, and take the optimization target or the value range of the optimization target corresponding to the radiator response surface model as the optimal design parameter combination. As in this implementation, the optimal parameters obtained are f / a=0.8, w / b=1.347, L=24cm, the arrangement of the bionic shark skin tooth fins is staggered mode 1, and the inlet and outlet layout mode selects a single inlet and double outlet layout mode. The corresponding predicted values ​​of T_ave, ΔP and TP are shown in Table 4.

[0075] In order to further ensure the accuracy of the optimization results, the predicted results are compared with the calculated results. Table 5 lists the relative deviations between the two. The relative errors of T_ave, ΔP and TP are 0.33%, 0.24% and 0.99%, respectively, which are all less than 3%. Therefore, this method has high-precision reliability.

[0076] Table 4 Optimal parameter combinations

[0077] f / a w / b L T_ave ΔP TP Predicted value 0.8 1.347 24 34.812 5.345 5.064

[0078] Table 5 Error analysis of optimization results

[0079]

[0080] The optimized fin radiator has better temperature uniformity, lower pump power consumption, and improves the overall performance of the radiator.

[0081] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.

Claims

1. A multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD, characterized in that: The steps include: S1. Establish a physical model of the liquid cooling radiator, determine the optimization target and evaluation index according to the structural characteristics of the liquid cooling radiator; take the structural parameters of the radiator as the optimization target; take the average temperature, pressure drop, and comprehensive evaluation index of the radiator as the evaluation index; S2. Using CFD technology combined with evaluation indicators, single factor analysis is performed on the determined multiple optimization targets to determine the optimization target or the value range of the optimization target; S3, generating several response surface analysis test plans based on the optimization target or the value range of the optimization target determined in S2; S4, based on several response surface analysis test schemes generated in S3, respectively establish the radiator response surface models corresponding to the schemes; S5. numerically simulate each established radiator response surface model, and calculate the evaluation index corresponding to each radiator response surface model; S6. Verify the linear regression results of the radiator response surface model; S7. Determine the optimal radiator response surface model based on the response surface analysis and linear regression results, and set the optimization target or the value range of the optimization target corresponding to the radiator response surface model as the optimal design parameter combination.

2. The multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: The structural parameters of the radiator include the inlet and outlet design, size, and internal structure design of the radiator.

3. The multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: If the radiator is a fin-type liquid-cooled radiator, its internal structure is designed with fin arrangement, fin spacing ratio f / a and w / b, inlet and outlet layout, and outlet spacing L; the front and rear of the fins are divided inside the radiator according to the coolant flow direction, then f is the spacing between adjacent fins on the left and right, w is the spacing between adjacent fins on the front and back, a is the span length of the fin, and b is the chord length.

4. The multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD according to claim 3, characterized in that: Fin arrangement methods include equal spacing arrangement, variable spacing arrangement, array alignment arrangement, and staggered arrangement.

5. The multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD according to claim 3, characterized in that: The import and export arrangements include single inlet and single outlet, single inlet and multiple outlets, multiple inlets and single outlet, and multiple inlets and multiple outlets.

6. The multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: The method of single factor analysis in S2 is: for the determined multiple optimization objectives, one of the optimization objectives is selected as a variable, and the other optimization objectives are fixed; the evaluation indicators corresponding to different variables are calculated respectively, and the optimal variable is determined based on the evaluation indicators.

7. The multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: The average temperature is expressed as: Where T is the surface temperature of the radiator, A W is the surface area of ​​the radiator, and A is the surface of the radiator.

8. The multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: The pressure drop is expressed as: ΔP=ΔP in -ΔP out Among them, ΔP in is the inlet gauge pressure, ΔP out is the outlet gauge pressure.

9. The multi-objective optimization method for a liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: The comprehensive evaluation index is expressed as: TP=No / f 1 / 3 Where Nu and f are the average Nusselt coefficient and the average friction coefficient, respectively.

10. The multi-objective optimization method for liquid cooling radiator based on response surface analysis and CFD according to claim 1, characterized in that: Several response surface analysis experimental schemes were obtained using Design-Experiment software.

Images