Optimization design of slotted fin by numerical simulation coupled with genetic algorithm
Yu Wang,Ya-Ling He,Dan-Hua Mei,Wen-Quan Tao
DOI: https://doi.org/10.1016/j.apenergy.2011.05.030
IF: 11.2
2011-01-01
Applied Energy
Abstract:Using a novel method that couples genetic algorithm (GA) with numerical simulation, the geometric configuration for a two-dimensional slotted fin has been optimized in this paper. The objective of optimization is to maximize the heat transfer capacity of slotted fin, and minimize the pressure drop penalty of fluid flow through the fin. The key of this method is the fitness function of GA, which were ( j / j 0 )/( f / f 0 ) and j / j 0 . In this complex multiparameter problem, the numerical simulation is a crucial step to calculate the Colburn factor j and friction factor f . The results showed that for two-dimensional slotted fin considered, the j factor is increased by 229.22%, the f factor is increased by 196.30%, and the j / f ratio was increased by 11.11% at Re = 500 based on optimal integrated performance ( j / j 0 )/( f / f 0 ); the j factor is increased by 479.08% at Re = 500 based on optimal heat exchange capacity j / j 0 . The feasibility of optimal designs was verified by the field synergy principle. Keywords Genetic algorithm Optimization Computational fluid dynamics (CFD) Heat exchanger Heat transfer Field synergy principle Nomenclature a thermal diffusivity, m 2 s −1 A heat transfer surface area, m 2 A c minimum flow area, m 2 c p fluid specific heat, J kg −1 K −1 D h hydraulic diameter, m f friction factor F fitness function h heat transfer coefficient, W m −2 K −1 H i strip height, mm j Colburn factor L flow length, mm L i strip width, mm N population quantity Nu Nusselt number p pressure, Pa Δ p pressure drop, Pa P fin pitch, mm P c the crossover probability P m the mutation probability Pr Prandtl number Q heat transfer rate, W Re Reynolds number St Stanton number T temperature, K Δ T logarithmic mean temperature difference, K u , v x , y velocity components, ms −1 x , y Cartesian coordinates Greek symbols δ fin thickness, mm θ the local intersection angle, deg λ thermal conductivity, W m −1 K −1 μ dynamic viscosity, Pa s ν kinematic viscosity, m 2 s −1 ρ fluid density, kg m −3 Subscripts b based on bulk temperature in tube inlet m mean or average value out tube outlet w will 1 Introduction Fin-and-tube heat exchangers are widely used in various engineering fields such as chemical process, air-conditioning, refrigeration, electronics cooling, and process industry. In order to reduce the energy consumption and benefit the economy, high-performance heat transfer components for thermodynamic process are becoming more and more popular for many applications [1] . The total thermal resistance for such kind of heat exchangers is comprised of three parts: the air-side convective resistance, the wall conductive resistance, and the liquid-side convective resistance. The heat transfer coefficient on the air-side is typically low due to the thermal conductivity of the air and relatively low frontal velocity. The use of enhanced fin surface is the most effective way to improve the overall performance of the fin-and-tube heat exchanger [2,3] . Owing to the requirement of smaller volume, low cost, quiet, and high efficiency in operation, different new types of enhanced fins, such as the wavy fin, the louvered fin and the slit fin, were developed. Experimental studies for these different types of fin have been extensively performed. Large number of investigations shows that slotted fin with protruding strips which are parallel to the base sheet has better performance than the wavy fin and the louvered fin [4,5] . This type of fin geometry was first studied by Nakayama and Xu [6] , who reported that its heat transfer coefficient can be 78% higher than that of the plain fin at an air velocity of 3 m/s. Hiroaki et al. [7] indicated that the heat exchanger with slotted fins can reduce the volume by 1/3 compared to that with plain fin. Recently, Wang et al. [8–10] conducted comprehensive experimental investigations for two types of slit fins with different protruding directions, and the experimental correlations of the heat transfer and flow friction were also proposed. Besides the studies on the heat transfer and flow characteristics of the entire fins, there are some investigations focused on the strips of the slit fin. Yun and Lee [11] analyzed the effects of various design parameters of the strip on the heat transfer and pressure drop characteristics of the slit fin heat exchangers, and presented the optimum value for each parameter. Kang and Kim [12] experimentally studied the effect of strip location on the heat transfer and pressure drop, and found that the slit fin with all the strips mainly positioned in the rear part has the better performance than that of the slit fin with all the strips in the front part. Qu et al. [13] numerically validated this finding and explained it from the viewpoint of field synergy principle, which was first proposed by Guo et al. [14,15] for parabolic fluid flow and heat transfer. Later, Tao et al. [16,17] extended this idea from parabolic flow to elliptic flow. Based on field synergy principle, Cheng et al. [18] proposed a new slotted fin with strips on the fin surface abiding by the rule of “front coarse and rear dense” along the flow direction. The numerical results show that the j factor of the new fin is about 9% higher than the fin with all the same number of strips in the front part. In recent years, a few academics had applied the genetic algorithm to heat transfer fields. The heat recovery steam generator [19] , the refrigerant circuit [20] were designed to optimize by genetic algorithm. But these optimization designs by genetic algorithm method cannot used to design new heat exchanger because they were typically based on empirical formula. In addition, genetic algorithm is often used to fit correlations based on experimental study results [21] . As previously mentioned, most of the optimization problem of slotted fin was solved by experimental method as yet. The aim of this study is to obtain the optimized design of slotted fin base on different performance evaluations by genetic algorithm coupled with numerical simulation. It is believed that this study not only could be used to optimize the slotted fin, but also provide a new method to design multifarious heat transfer components. 2 Genetic algorithm model design Two targets of the slotted fin heat exchanger optimization design are considered: one is to obtain the minimum pressure drop penalty, and the other is to obtain the maximum heat exchange capacity. Strip height is the structural parameters that have an important influence on heat transfer and fluid flow characteristics of the heat exchanger. Therefore, the heights of each strip were chosen as the parameters of the optimization problems. As the strips are positioned symmetrically alone the flow direction, there are seven independent parameters. Since each height can be changed continuously, this optimization problem has very tremendous parameter domain, which could be perfectly solved by the genetic algorithm. The genetic algorithm (GA) uses a population of several individuals to perform the optimization by simulating the genetic and evolution mechanism of biology [22] . It generally contains six steps: coding/decoding, population initialization, evaluation, selection (or reproduction), crossover, and mutation. The individuals are represented with strings in a binary format, and the initial individuals are randomly generated from the search space. Each individual is evaluated and assigned with a fitness value determined by the fitness function. The higher fitness value individual has the higher chance to survive to the next generation during the selection process. Some individuals in the new generation are generated from individuals selected at a given probability by using crossover and mutation operations. Crossover operation generates new individuals by exchanging the data between two randomly chosen individuals, and mutation operation generates new individuals by randomly changing data of one randomly chosen individual. The individuals in the new generation are evaluated and the process is repeated until the given maximum number of generations is reached or convergence is reached. The program flow chart of overall process is shown in Fig. 1 . The height of each strip was chosen as the chromosomes and encoded in the regions from −7.0 × 10 −4 to 7.0 × 10 −4 . The number of the chromosomes that keep in each generation is denoted by N = 64. All individuals were chosen as the initial generation population and their corresponding fitness values were then calculated. Two kinds of performance evaluations have been provided as the fitness function of GA. They are defined as (1) F 1 = ( j / j 0 ) / ( f / f 0 ) (2) F 2 = j / j 0 The better fitness of chromosomes, the higher chance of survival. These surviving chromosomes then reproduced new populations by selection, crossover and mutation operations. The probability of performing a crossover in the GA methods is denoted by P c = 0.5. The probability of performing mutation in the GA method is denoted by P m = 0.02. 3 Numerical simulation model description Since the heat transfer through fin surface accounts for about 90% of the total heat transfer, the computational model could be simplified into two-dimensional. In addition, the slotted fin was chosen as symmetrical because of processing and installation requirements. A schematic view of a two-dimensional array of slotted fins, which were positioned periodically, is shown in Fig. 2 . The pitch of each fin P is 1.6 mm, the flow length of each period L is 18.0 mm, and fin thickness δ is 0.1 mm. In this study, we define x , y as primary and secondary streamwise coordinates, respectively, and y stands for the fin pitch direction. Due to periodicity, the region sketched by the dashed lines in Fig. 2 is selected as the computational domain. The centric surfaces of the neighboring two pitches are selected as the upper and lower boundaries of the computational domain, which are set as y direction periodic boundary conditions. The face of upstream and downstream are set as x direction periodic boundary conditions. As seen in Fig. 3 , the scheme of slit is symmetrical along the flow direction, H i and L i stands for the height and width of each slit, respectively, and H i is just the parameters that require optimizing in this paper. L i is evenly arranged by 10.0 mm. Due to the relatively high thermal conductivity of the fin, it is assumed that the fin surface temperature is a constant. The air inlet temperature is 303.15 K and the fin temperature is 323.15 K. 4 Mathematical formulation and numerical method 4.1 Governing equations and boundary conditions Due to the small air temperature difference over the fin length, the air is considered as an incompressible fluid with constant thermophysical properties. The flow in the computational domain is assumed to be two-dimensional, laminar, steady state, and periodically fully developed. In addition, the body force and the viscous dissipation term are neglected. The governing equations including continuity, momentum and energy equation for the fluid domain can be expressed as follows. Continuity equation: (3) ∂ u ∂ x + ∂ v ∂ y = 0 Momentum equations: (4) u ∂ u ∂ x + v ∂ u ∂ y = - ∂ p ∂ x + ν ∂ 2 u ∂ x 2 + ∂ 2 u ∂ y 2 (5) u ∂ v ∂ x + v ∂ v ∂ y = - ∂ p ∂ y + ν ∂ 2 v ∂ x 2 + ∂ 2 v ∂ y 2 Energy equation: (6) u ∂ T ∂ x + v ∂ T ∂ y = a ∂ 2 T ∂ x 2 + ∂ 2 T ∂ y 2 Because the governing equations are elliptic in spatial coordinates, the boundary conditions are required for all boundaries of the computational domain. The temperature of the slotted fin is higher than that of the inlet air. The required boundary conditions are described as follows. At the upper and lower boundaries: periodic boundary conditions. (7) u ( x , 0 ) = u ( x , P ) (8) v ( x , 0 ) = v ( x , P ) (9) T ( x , 0 ) = T ( x , P ) At the front and back boundaries: periodic boundary conditions. (10) u ( 0 , y ) = u ( L , y ) (11) v ( 0 , y ) = v ( L , y ) (12) T ( 0 , y ) - T w T m ( 0 ) - T w = T ( L , y ) - T w T m ( L ) - T w Fin surface region: no-slip conditions and constant temperature T w are defined. (13) u wall = 0 (14) v wall = 0 (15) T wall = T w 4.2 Numerical methods The governing equations are discretized by the finite volume method. The SIMPLE algorithm [23] is utilized to deal with the coupling of pressure and velocity. The fluid–solid conjugated heat transfer problem is solved by full-field computation method. The solid in computational domain is regarded as a special fluid with an infinite viscosity. The detailed computational method of conjugated heat transfer can be found in Tao [23,24] and Patankar [25] . A very large value of the thermal conductivity is assigned to the fin region to guarantee the fin temperature to be uniform. The computation domain is mashed by structured grid as shown in Fig. 4 . The height of each strip H i in the figure are 0.6 mm, 0.2 mm, −0.2 mm, −0.6 mm, −0.2 mm, 0.2 mm, and 0.6 mm, and this configuration will be used to check the grid independence later. The convergence criterion for the continuity and momentum equations is that the maximum mass residual of the cells divided by the inlet mass flux is less than 1.0 × 10 −6 , and the criterion for the energy equation is that relative difference between the two Nusselt numbers in the successive iterations is less than 5.0 × 10 −7 . The value of each H i was produced by GA to form a kind of strip configuration. The fitness function of each configuration was simulated by the SIMPLE algorithm. 4.3 Parameter definitions To improve the physical understanding, some characteristic and non-dimensional parameters need to be defined. The Reynolds number is defined as (16) Re = ρ u m D h μ where u m is the mean velocity at the minimum flow cross-sectional area A c , and D h is the characteristic diameter that is determined by (17) D h = 4 A c P The Colburn factor j is used to describe the heat transfer performance: (18) j = St · Pr 2 / 3 = Nu Re · Pr 1 / 3 where the Nusselt number Nu is defined as follows. (19) Nu = hD h λ The heat transfer coefficient h is defined in terms of the heat transfer rate Q and the logarithmic mean temperature difference: (20) h = Q A Δ T where the overall heat energy Q added to the fluid can be calculated by measuring inlet and outlet temperatures of air flow, T in and T out . (21) Q = m ˙ c p T ¯ out - T ¯ in where m ˙ is the mass flow rate obtained by (22) m ˙ = ρ u m A c The logarithmic mean temperature difference is expressed as (23) Δ T = T ¯ max - T ¯ min ln ( T ¯ max / T ¯ min ) (24) where T max = max ( T ¯ w - T ¯ in , T ¯ w - T ¯ out ) (25) T min = min T ¯ w - T ¯ in , T ¯ w - T ¯ out The mean temperature of a cross section are defined as (26) T ¯ = ∫ A uTdA ∫ A udA The friction factor f is used to describe the pressure loss characteristics. (27) f = p ¯ in - p ¯ out L D h ρ u m 2 2 Finally, the mean pressure of a cross section are defined as (28) p ¯ = ∫ A updA ∫ A udA 5 Grid independence and model validation 5.1 Grid independence The investigation of grid independence is necessary to ensure the accuracy and validity of the numerical results. The grid independent study is performed on the slotted fin with the given protruding scheme, which was mentioned before, for Re = 500. In order to investigate the influence of the grid density on the computational results, three sets of grid number are studied: 182 × 34, 200 × 37, and 218 × 41. The results of the three sets with different grid number are tabulated in Table 1 . The relative error of the Nusselt number and the friction factor between three types of the grid number is less than 1% and 2%, respectively. As these different are very small, to save computer resource and keep a balance between computational cost and accuracy, the adopted grid number in the computational domain is about 182 × 34. 5.2 Model verification In order to validate the computational model and numerical method, simulation is carried out at the slotted fin with H i = 0.0 mm ( i = 1–7). Thus, the problem was transformed into solving the problem of infinite parallel plates, which is the geometry amenable to the simplest mathematical treatment. The Reynolds number, Re, ranges from 100 to 1500, and the corresponding frontal air velocity ranges from about 0.5 m/s to 8.0 m/s. As the laminar flow and heat transfer for the parallel plates have been analyzed in great detail, the classical result was used to validate the computational model and numerical method. The fully developed laminar velocity profile and friction factor for a plate spacing of 2 b with coordinate axes at the center are given by Rohsenow and Choi [26] and McCuen et al. [27] as (29) u = 1 2 c 1 y 2 - b 2 (30) u m = - 1 3 c 1 b 2 (31) fRe = 24 The comparisons between friction factors in the present simulation and the classical results above are shown in Fig. 5 . It is seen that the deviation is within 7.5% for 95% of the numerical result. The fully developed Nusselt number with equal wall temperature is the constant and is designated as [28] (32) Nu = 7.54070087 The comparisons between Nusselt numbers in the present simulation and the above classical results are shown in Fig. 5 . It is seen that the maximum deviation is within 1.5%. 6 Numerical results and discussions 6.1 Genetic algorithm results In the present work, genetic algorithm (GA) based approach to solve the optimization problem of slotted fin heat exchanger with the different strip height is conducted. Based on the aforementioned optimization method, the results are shown as follow. After about 100 generation and optimization tracing cycles, the globally optimum design results of configurations for slit fin heat exchangers were solved. The optimization tracing figures of the genetic algorithm based on different fitness functions are shown in Fig. 6 . It can be seen that convergence is reached after about fifty cycles or less. The optimal structures of slotted fin heat exchanger based on two different fitness functions are shown in Fig. 7 . For convenience, Baseline Case refers to the plate fin, while Case 1 and Case 2 are used to represent the results of optimization by genetic algorithm based on fitness functions F 1 and F 2 , respectively. The optimization results of each case are tabulated in Table 2 . 6.2 Fluid flow and heat transfer performances In order to study the heat transfer performances and flow characteristics for optimally designed slotted fin heat exchangers based on different fitness function, a comparative parametric study is performed. The protruding strips are symmetrically disposed alone the flow direction. The Reynolds number, Re , based on the hydraulic diameter is about 500. Different configurations for slotted fin heat exchangers were presented above. Fig. 8 presents the local velocity distribution for the baseline case and two different slotted fin case at Re = 500. It can be seen from Fig. 8 a that a boundary layer region alone the fin is formed. The boundary layer flow is separated from the main stream, and the fluid in this region is thermally isolated. Consequently, a poor heat transfer zone is developed. By comparing different fin patterns in Fig. 8 , it can be seen that the velocity distribution and the wake structure for the heat transfer enhancement case are totally different from those for baseline case. Fig. 9 shows the dimensionless local temperature distribution on the cross section for the baseline and two different fin patterns ( T f − T in )/( T w − T in ) at Re = 500. By comparing three different fin patterns, the mass average temperature at the inlet region is identical for the baseline case and the enhanced cases. However, the heat transfer enhancement is notable in the vicinity of the downstream region because of the protruding strips. As the performance evaluation that was used as fitness function not only the capability of heat transfer, but also the pressure drop penalty is considered. The temperature behind the strips for the Case 1 is distinctly higher than that in the corresponding region for the baseline case and lower then that for the Case 2. In comparison with the baseline case at Re = 500, the Colburn factor j of the Case 1 is increased by 229.22%, and the friction factor f is increased by 196.30%. The overall performance of the slotted fin heat exchanger with different strip configuration is evaluated using the criterion of the flow area goodness factor j / f . For Case 1, the j / f ratio was increased by 11.11% over the baseline case. This result indicated that the Case 1 strip configuration can significantly enhance the heat transfer with moderate pressure loss penalty. For the enhanced configuration Case 2, the performance evaluation that was used as fitness function only the capability of heat transfer is considered, no attention is paid to the flow friction. Hence, the optimization result only pursues higher heat transfer enhancement. The temperature at downstream region for the Case 2 is distinctly higher than that in the corresponding region for the other two cases. Comparing with the baseline case, the Colburn factor j of the Case 2 is increased by 479.08%, but the friction factor f of is increased by 555.71% at Re = 500. The results indicate that this configuration can significantly enhance the heat transfer at the expense of considerable pressure loss penalty. 6.3 Fundamental mechanism for heat transfer enhancement As mentioned above, from the traditional perspective, vortex generators can enhance convective heat transfer because they can disrupt the boundary layer, intensify the mixing, and increase the disturbance in the channel. All these mechanisms contribute to heat transfer augmentation. However, this is insufficient to reveal the fundamental mechanism for heat transfer enhancement. According to the field synergy principle (FSP) proposed by Guo et al. [14,15] , and proved by Tao and co-workers [16,17] , the existing single phase convective heat transfer enhancement methods can be unified by FSP. For brevity, the details about FSP can be found in relevant Refs. [14–17] and will not be repeated here. The local intersection angle between the velocity vector and the temperature gradient at a grid node is defined as (33) θ = cos - 1 U · ∇ T | U | | ∇ T | = cos - 1 u ∂ T ∂ x + v ∂ T ∂ y | U | | grad T | The average intersection angle of the computation domain can be obtained by using numerical integration: (34) θ m = ∑ i , j θ i , j Δ x i Δ y j ∑ i , j Δ x i Δ y j where the subscripts i and j refer to the control volumes of the fluid. The average intersection angles for the baseline configuration and enhanced configurations are presented in Table 3 . From the table we can see that at the same Reynolds number the average intersection angle θ m for the enhanced configurations is smaller than that for the baseline configuration, which indicates that the synergy for the enhanced configurations is better than that for the baseline configuration. For the same Reynolds number, the average intersection angle of the baseline configuration is the largest, while that of the slotted fin Case 2 is the smallest, which is in accordance with their corresponding heat transfer performance. In order to have a more intuitive understanding of the field synergy principle, Fig. 10 presents the streamlines and the isothermal distributions for baseline configuration and enhanced configurations at Re = 500. For the baseline case, it can be seen clearly in Fig. 10 a that the isotherms are almost paralleled to the local streamlines; this indicates that the direction of temperature gradient is almost perpendicular to the velocity vector. This implies a poor synergy between the velocity field and the temperature field. It appears that the bad heat transfer regions are the same regions of poor synergy, which is consistent with the field synergy principle. Due to the presence of the enhanced Case 1, the velocity field and the temperature field are modified. It can be seen clearly from Fig. 10 b that the synergy between the velocity field and the temperature field for Case 1 is significantly improved, especially in the downstream region. For the enhanced Case 2, it can be seen from Fig. 10 c that the synergy between the velocity and the temperature gradient is further improved in comparison with Case 1. The better synergy is mainly due to the two-side arrangement of the strips. For Case 1, no strip side was in the weak region that is associated with the low heat transfer coefficient. However, for Case 2, the upstream fluid will directly encounter with the slotted fin in all dimensions, and it will augment the heat transfer. Consequently, the heat transfer of Case 2 is better than that of Case 1. 6.4 Future work Heat exchangers play an important role in traditional energy process as well as the new energy techniques and applications. Much more work is required for how to optimize design a heat exchanger with better comprehensive performance. Rigorous and generalized procedure for genetic algorithm coupled with numerical simulation is still in the early stages of development. New computational algorithms are needed to decrease the computational time and increase confidence in the predictive capability in entire model simulations. New optimization algorithms are needed for the actual complex model. In order to popularize this new algorithm and improvements, three-dimensional model are needed to consider by three-dimensional optimization program. In addition, accurate heat transfer and pressure drop date of complex model will be obtained by mixed-language programming of genetic algorithm and commercial computational fluid dynamics code. In order to maximize the efficiency of this new algorithm and improvements, multi-node parallel process technology must be carried out to save computation time as well as the capital cost and operating cost. Regarding the practical application, further developments in genetic algorithm coupled with numerical simulation method are required to present complex three-dimensional model. Ensure the accuracy and reduce the computation cycle and time of this novel optimization algorithms are also important and need to be investigated. 7 Conclusions This paper presents an intelligent design method to optimize the configuration of a two-dimensional slotted fin heat exchanger. Two kinds of strip configurations were optimized by GA based on two different fitness functions, and the comparison of the fluid flow structures and heat transfer characteristics among three cases is explored. The major conclusions are drawn as follows. (1) The genetic algorithm coupled with numerical analysis was used to optimize design a two-dimensional slotted fin heat exchanger. This method is very appropriate for optimizing the nonlinear, multi-parameter, enormous parameter space, complex problems. Therefore, this method could be used to optimal design multifarious heat transfer components. (2) For Case 1, it was optimized based on comprehensive evaluation of F 1 . Comparing with the baseline case at Re = 500, the Colburn factor j is increased by 229.22%, and the friction factor f is increased by 196.30%. It is worth noting that the increase of Colburn factor j is larger then that of friction factor f and the j / f ratio was increased by 11.11% over the baseline case. (3) The enhanced Case 2 was optimized at Re = 500 based on fitness function F 2 , which only considers the heat transfer performance. Hence, this optimal design case has superduper heat transfer capacity. By comparing with the baseline case at Re = 500, the Colburn factor j is increased by 479.08%. (4) For the Reynolds number 500, the average intersection angles θ m decreases in the order of the baseline case, Case 1, and Case 2, and the heat transfer performance increases in the same order. The smaller intersection angle θ m means the better synergy between the velocity and the temperature gradient. The better synergy causes higher heat transfer performance for the heat exchanger. Acknowledgements This work is supported by the Key Project of National Natural Science Foundation of China (No. 50736005), the Joint Funds of NSFC-Guangdong of National Natural Science Foundation of China (No. U0934005), and the National Basic Research Program of China (973 Program) (2011CB710702). References [1] Y.G. Lei Y.L. He R. Li Y.F. Gao Effects of baffle inclination angle on flow and heat transfer of a heat exchanger with helical baffles Chem Eng Process 47 12 2008 2336 2345 [2] Y.L. He W.Q. Tao F.Q. Song W. Zhang Three-dimensional numerical study of heat transfer characteristics of plain plate fin-and-tube heat exchangers from view point of field synergy principle Int J Heat Fluid Flow 26 3 2005 459 473 [3] Y.B. Tao Y.L. He J. Huang Z.G. Wu W.Q. Tao Numerical study of local heat transfer coefficient and fin efficiency of wavy fin-and-tube heat exchangers Int J Therm Sci 46 8 2007 768 778 [4] H.J. Kang W. Li H.J. Li R.C. Xin W.Q. Tao Experimental study on heat transfer and pressure drop characteristics of four types of plate fin-and-tube heat exchanger surfaces J Therm Sci 3 1 1994 34 42 [5] J.Y. Yun K.S. Lee Investigation of heat transfer characteristics on various kinds of fin-and-tube heat exchangers with interrupted surfaces Int J Heat Mass Transfer 42 1999 2375 2385 [6] Nakayama W, Xu LP. Enhanced fins for air-cooled heat exchangers-heat transfer and friction factor corrections. In: Proceedings of the 1983 ASME–JSME thermal engineering conference, vol. 1; 1983. p. 495–502. [7] K. Hiroaki I. Shinichi A. Osamu K. Osao High-efficiency heat exchanger Natl Tech Rep 35 6 1989 653 661 [8] C.C. Wang W.H. Tao C.J. Chang An investigation of the air side performance of the slit fin-and-tube heat exchangers Int J Refrig 22 1999 595 603 [9] Y.J. Du C.C. Wang An experimental study of the air-side performance of the super slit fin-and-tube heat exchangers Int J Heat Mass Transfer 43 2000 4475 4482 [10] C.C. Wang W.S. Lee W.J. Sheu A comparatives study of compact enhanced fin-and-tube heat exchangers Int J Heat Mass Transfer 44 2001 3565 3573 [11] J.Y. Yun K.S. Lee Influence of design parameters on the heat transfer and flow friction characteristics of the heat exchanger with slit fins Int J Heat Mass Transfer 43 2000 2529 2539 [12] H.C. Kang M.H. Kim Effect of strip location on the air-side pressure drop and heat transfer in strip fin-and-tube heat exchanger Int J Refrig 22 1998 302 312 [13] Z.G. Qu W.Q. Tao Y.L. He 3D numerical simulation on laminar heat transfer and fluid flow characteristics of strip fin surface with X-arrangement of strips J Heat Trans-T ASME 126 2004 697 707 [14] Z.Y. Guo D.Y. Li B.X. Wang A novel concept for convective heat transfer enhancement Int J Heat Mass Transfer 41 1998 2221 2225 [15] Wang S, Li ZX, Guo ZY. Novel concept and devices of heat transfer augmentation. In: Proceedings of 11th international conference of heat transfer, vol. 5. London: Taylor & Francis; 1998. p. 405–8. [16] W.Q. Tao Z.Y. Guo B.X. Wang Field synergy principle for enhancing convective heat transfer: its extension and numerical verifications Int J Heat Mass Transfer 45 2002 3849 3856 [17] W.Q. Tao Y.L. He Q.W. Wang Z.G. Qu F.Q. Song A unified analysis on enhancing single phase convective heat transfer with field synergy principle Int J Heat Mass Transfer 45 24 2002 4871 4879 [18] Y.P. Cheng Z.G. Qu W.Q. Tao Y.L. He Numerical design of efficients lotted fin surface based on the field synergy principle Numer Heat Transfer A-Appl 45 6 2004 517 538 [19] M. Mohagheghi J. Shayegan Thermodynamic optimization of design variables and heat exchangers layout in HRSGs for CCGT, using genetic algorithm Appl Therm Eng 29 2009 290 299 [20] Z.G. Wu G.L. Ding K.J. Wang M. Fukaya Application of a genetic algorithm to optimize the refrigerant circuit of fin-and-tube heat exchangers for maximum heat transfer or shortest tube Int J Therm Sci 47 2008 985 997 [21] Q.W. Wang D.J. Zhang G.N. Xie experimental study and genetic algorithm based correlation on pressure drop and heat transfer performances of a cross-corrugated primary surface heat exchanger J Heat Transfer-T ASME 131 2009 061802 [22] S. Sanaye H. Hajabdollahi Thermal-economic multi-objective optimization of plate fin heat exchanger using genetic algorithm Appl Energy 87 6 2010 1893 1902 [23] W.Q. Tao Numerical heat transfer 2nd ed. 2001 Xi’an Jiaotong University Xi’an [24] W.Q. Tao Recent advances in computational heat transfer 2000 Science Press Beijing [25] S.V. Patankar Numerical heat transfer and fluid flow 1980 McGraw-Hill New York [26] W.M. Rhosenow H.Y. Choi Heat, mass and momentum transfer 1961 Prentive-Hill Englewood Cliffs, New Jersey [27] McCuen PA, Kays WM, Reynolds WC. Heat transfer with laminar and turbulent flow between parallel planes with constant and variable wall temperature and heat flux. Rep. No. AHT-3. Department of Mechanical Engineering, Stanford University, Stanford, California; 1962. [28] Shah RK, London AL. Laminar flow forced convection in ducts. In: Hartnett JP, Irvine TF, editors. Advances in heat transfer, suppl. 1. New York: Academic Press; 1978.