The 6th International Supercritical CO2 Power Cycles Symposium March 27-29, 218, Pittsburgh, Pennsylvania Channel Structure Inluence on the Thermal-Hydraulic Perormance o Zigzag PCHE Yichao Gao Wenkai Xia Yaowu Huo Yiping Dai Master Student Doctor Student Doctor Student Proessor Xi an Jiaotong University Xi an Jiaotong University Xi an Jiaotong University Xi an Jiaotong University Shaanxi, China Shaanxi, China Shaanxi, China Shaanxi, China Yichao Gao 211-215, B.S. in Power Engineering, Xi an Jiaotong University, Shaanxi, China 215-Present, M.S. in Power Engineering, Xi an Jiaotong University, Shaanxi, China ABSTRACT Printed Circuit Heat Exchanger (PCHE) is a very promising lat plate heat exchanger. Based on the PCHE model o Tokyo Institute o Technology, CFD method was used to measure the two in angles which were 32.5 and 4. o the zigzag channels, the simulation shows a good agreement with the experiment results in local heat transer, Nusselt number and riction actor. The channel structure inluence on the thermal-hydraulic perormance o PCHE was studied, including 1.-6.mm channel width, 5-6 in angle, and six types o in length, which were 27mm 2, 18mm 3, 9mm 6, 6mm 9, 3mm 18 and 2mm 27. The heat transer and pressure drop increases with the decrease o channel width and the increase o in angle. The 1.-2.mm channel width and 2-45 in angle are recommended. It indicates that the in length is not as small as possible. 3mm in length is the turning point and is avorable or its optimal heat transer and relatively limited pressure drop. Keywords: supercritical carbon dioxide; printed circuit heat exchanger; channel width; in angle; in length 1 Introduction Printed Circuit Heat Exchanger (PCHE) is a compact heat exchanger with excellent heat transer eectiveness [1], which was released in Australia in 198. In 1985, PCHE was irst used by Heatric in the rerigeration cycle in UK, representing the beginning o the commercialization o PCHE. The luid channels o the PCHE are ormed by pochemical etching process on a metal plate. The conventional cross-sectional shape o the channel is semicircle with the diameter o 1.-2.mm, and 1
dierent plates are stacked together by diusion bonding to assemble the heat exchanger [2], as shown in Fig.1 [3]. Figure.1. PCHE Channel Cross Section and Core Structure PCHE can meet the heat transer process o high pressure, high eectiveness, less leakage, compact structure, and so on. The maximum pressure that it can withstand is 6MPa, while the maximum temperature is 9, the eectiveness is over.9, and some are even as high as.98. At the same heat load and pressure drop, the volume o PCHE is 1/6-1/4 o the conventional shell and tube heat exchanger, and the average unit mass heat load reaches to 2kg/MW, and the heat transer area density is up to 25m 2 /m 3. Due to its compactness and good heat transer properties, PCHE is being used in a wide range o industrial ields, such as chemical processes, uel processing, aviation, aerospace and nuclear energy, which is a very promising candidate or the intermediate heat exchanger (IHX) in high temperature gas-cooled reactors (HTGRs) and sodium-cooled ast reactors (SFRs). There are our types o PCHE according to the shape and layout o dierent low channels: lat, zigzag, S- shaped and airoil. Conventional PCHE low channels are continuous semicircular zigzag corrugated channels, each o which can be used as a channel with many bends, where eddy currents, counter current, and vortices at corners improve luid heat transer perormance and also increase the pressure drop. Thereore, some researchers have been interested in studying the internal heat transer and low properties in PCHE with dierent channel structures. Figure.2. Dierent PCHE Flow Channels Nikitin et al. [4] conducted an experiment on the low and heat transer perormance o supercritical carbon dioxide in the conventional zigzag corrugated channel PCHE. The results show that the local heat transer coeicient and pressure drop coeicient o the luid is a unction o the Reynolds number (Re). Ngo et al. [5] designed a sinusoidal S-ribbed corrugated structure that simpliied the heat transer between 2
the carbon dioxide and water sides to a single heat transer unit and made numerical calculation. The results show that the volume o the PCHE can be reduced to 1/3 compared with that o the conventional zigzag corrugated channel, the pressure drop on the carbon dioxide and water side can be reduced by 37% and 9%, respectively. Kim et al [6]. studied the heat transer and pressure drop o supercritical carbon dioxide in a new PCHE, o which the in structure is the NACA2 streamlined wing. The results show that the new PCHE has the characteristics o large heat transer area and uniorm low. The total heat transer rate per unit volume is basically the same as the conventional zigzag PCHE, but the pressure drop is only 1/2 o the conventional PCHE. Tsuzuki et al [7]. proposed a PCHE with S-shaped in structure. By studying the inluence o the in length and in angle on the pressure drop and heat transer perormance, a more optimized low path structure was obtained. The pressure drop o the new channel structure is 2% o the conventional zigzag structure with the same heat transer eectiveness, and the pressure drop is reduced by reducing the vortex at the elbow and reverse lows. To sum up, it can be ound that improving PCHE channel structure and reducing pressure drop on the basis o improving heat transer are the research emphases, which is o great signiicance or improving the overall perormance o heat exchanger. In the experiments, it is very diicult to measure the local low parameters in the micro-scale channels. Thereore, CFD (Computational Fluid Dynamics) is widely used or doing research on PCHE. 2 Model Establishment The experiment model o TIT (Tokyo Institute o Technology) was used in this paper, as shown in Fig.3. The PCHE volume is 71 76 896mm 3, including 144 channels and 66 channels. The diameter o the and channel is 1.9mm and 1.8mm, respectively. The double banking coniguration is adopted to ensure the same velocity or and channels under the condition that the volume low rate o luid is about twice o that in luid. Figure.3. PCHE Cold and Hot Channels Arrangement 3
2.1 Model Simpliication In the experiments conducted by Nikitin and Ishizuka et al. [8], the in angle o the and channel are 32.5 and 4., making the channel 1% longer than the channel, as shown in Fig.4. In order to acilitate the boundary conditions setting o the PCHE, the length o the and channels are considered to be the same. The in angle o 32.5 and 4. or both and channels were established [9], as shown in Fig.4. Because o the channel arrangement in the PCHE, it is not necessary to build the entire experiment model. Instead, one PCHE unit composed o two channels and one channel is established with periodic boundary conditions, which can simulate the entire PCHE heat transer situation accurately, as illustrated in Fig.5. Figure.4. PCHE Channel Unit Structure Parameters Figure.5. Experiment Model Simpliication In addition, the channel length has also been simpliied rom 896mm to 54mm. With this simpliication, the inlet temperature o the luid needs to be adjusted rom 18 C to 221.8 C, which proves to be a good match with the experiment in the numerical study o Kim et al. [1], as shown in Fig.6. The actual range o PCHE unit studied in this paper is the part corresponding to the X-axis -.54m. 4
Figure.6. Kim Simulation o Cold and Hot Fluid Temperature Change along the Wall 2.2 Meshing Solidworks sotware was used to generate the PCHE model. Suraces From Edges and Fill module in the Ansys design modeler was used to generate the SCO 2 Fluid. Tetrahedral network was used to mesh the PCHE and the SCO 2 luid. For the boundary layer between the solid and luid, Inlation module was applied and the irst layer thickness o SCO 2 luid is.1mm. Five rows were created with a growth rate o 1.2, the total thickness o the boundary layer is.744mm, according to the experience o Kim et al [11]. Grid independent test was carried out and results are listed in Table 1. When the number o elements reaches to 16988, heat transer coeicient on both sides converged. Hence, the 16988 elements was chosen as a reerence grid considering the simulation time and accuracy. Table 1. PCHE Grid Independence Test Description Mesh Elements HTC in Cold Side Relative Deviation HTC in Hot Side Relative Deviation Coarse 37841 671.3-12.1% 336.5-11.3% Medium 586652 7154.37-6.1% 3459.75-7.1% Medium-Fine 824445 7347.5-3.6% 3431.81-7.9% Fine 16988 7542.77-1.1% 3647.83-2.1% Good 1235863 7622.85.% 3726.9.% Note: Where HTC is the heat transer coeicient 2.3 Boundary Conditions Post-processing was perormed on Ansys CFX. The structural material o PCHE is steel, which has the thermal conductivity o 16.2W/(m K), and the heat transer mode is thermal energy. CO2RK was chosen or SCO 2 with the pressure ranging rom 1MPa to 2MPa, temperature ranging rom 1K to 1K. The total energy equation was used or heat transer and the k-ε turbulence model is adopted which has good convergence and is widely used in engineering. Inlet temperature and pressure or and side are 28, 3.2MPa and 221.8, 1.5MPa, respectively. 5% turbulence intensity was applied. Mass low rate control was used or outlet conditions, which is rom 3kg/h to 39kg/h, with an increment o 15kg/h. The top and bottom, let and right suraces o PCHE are set as translational periodicity boundary condition and the ront and rear suraces are set as adiabatic boundary conditions. For the boundary layer between 5
PCHE and SCO 2 luid, general connection was used, as illustrated in Fig.7. The irst-order high-precision convergence mode was adopted and the convergence precision is 1-4. Figure.7. PCHE Meshing and Boundary Settings 3 Comparison o Simulation and Experiment Two key metrics or measuring PCHE perormance are heat transer and pressure drop. Local heat transer coeicient h and Nusselt number were used to measure the heat transer perormance. Nusselt number characterizes the ratio o the luid thermal conductivity to the convective thermal resistance. The larger Nusselt number indicates better heat transer perormance o the PCHE: h D av Nu h av and λ av represent the average convective heat transer coeicient and average thermal conductivity o the luid, respectively. av h Pressure drop was deined by the subtraction o inlet and outlet pressure: P P P in In addition, the riction actor can also be used to measure the pressure drop. The larger indicates greater low resistance: out P P D in out h ( ) 2 2 V L out out P in and P out are the PCHE channel inlet and outlet pressure, respectively. ρ out and V out are the SCO 2 outlet density and velocity. CFD simulation results are compared with the experiment results in this chapter. Some researchers give the empirical ormula or the local heat transer coeicient h, Nusselt Number and riction actor related to the Reynolds number, Prandtl number and channel structure, which are shown in the ollowing table. 6
Table 2. Empirical Formula o Local Heat Transer Coeicient, Nusselt Number and Friction Factor Researchers Local HTC or Nu Friction Factor Application Scope Nikitin h h 2.52 Re 5.49 Re.681.625 6 1.42 1 Re.4495 6 1.545 1 Re.9318 28 Re 58 62 Re 121 Ishizuka h.214 Re 44.16 6 2. 1 Re.467 6 2. 1 Re.123 24 Re 6 5 Re 13.629.317 Ngo Nu.1696 Re Pr.339 Re.3749 Re.158.154 35 Re 22.75 Pr 2.2 Nu.4 Re Pr 2b ( ) a.64 1/ 3.75 Note: Where a is the in length and b is the channel width. 4.8 Re 2b ( ) a.36 1.5 4 5 1 Re 1 3.1 Heat Transer Comparison in Cold and Hot Side Due to the same thermal-hydraulic perormance between the 32.5 and 4. in angle o the PCHE, 32.5 in angle was used or description. As illustrated in Fig.8, (a), (b), (c) and (d) represent the CFD results o the local heat transer coeicient and Nusselt number compared with the experiment o dierent researchers, respectively. The ranges o the existing correlations were extended up to Reynolds number 6. Because o the similar low velocities o and luids, while the luid density is much higher than that o luids, the range o Reynolds numbers o luids is broader than that o luids. For the local heat transer coeicient h, the CFD result is in good agreement with the experiment results o Nikitin and Ishizuka at low Reynolds numbers. With the increase o Reynolds number, the CFD result is basically consistent with the trend o Ishizuka experimental result, while the deviation with Nikitin experimental result increases gradually. Taking CFD results as a reerence, Nikitin's experimental results can be applied to low Reynolds numbers, while Ishizuka's experimental results is recommended at both low and high Reynolds numbers. For the Nusselt number, the CFD result agrees with that o Ngo at low Reynolds number, while at high Reynolds number, the dierence between Ngo and CFD result become obvious. [12] experimental results is applicable at high Reynolds numbers. With the increment o Reynolds number, CFD results and experiment results become more consistent. 7
12 1 CFD Nikitin Ishizuka 6 5 CFD Nikitin Ishizuka -1 h/w m -2 K 8 6 4-2 K -1 h/w m 4 3 2 2 1 1 2 3 4 5 6 Re (a) 5 1 15 2 Re (b) 35 3 CFD Ngo 175 15 25 125 2 1 Nu Nu CFD Ngo 15 75 1 5 5 25 1 2 3 4 5 6 5 1 15 2 Re Re (c) (d) Figure.8. Heat Transer Perormance at Fin Angle 32.5 in Cold and Hot Side 3.2 Pressure Drop Comparison in Cold and Hot Side As shown in Fig.9, it indicates that the CFD results are in good agreement at in angle 32.5 with the experiment results rom our researchers. The riction actor decreases gradually and approaches a stable value with the increase o the Reynolds number. For the experimental results o, the riction actor is larger compared with the CFD results. For Nikitin and Ishizuka, the experimental results are more credible when the Reynolds number is low, while at high Reynolds number, the riction actor calculated by the empirical ormula is or negative, which is possibly inaccurate..2.15 CFD Nikitin Ishizuka Ngo.25.2 CFD Nikitin Ishizuka Ngo.15.1.1.5.5. 1 2 3 4 5 6 Re. 5 1 15 2 Re Figure.9. Pressure Drop at Fin Angle 32.5 in Cold and Hot Side 8
3.3 Relative Deviation o Heat Transer and Pressure Drop The relative deviation o heat transer and riction actor are shown in Table 3 and Table 4. By comparing the CFD results with the experimental data, it is concluded that the relative deviation are acceptable. The accuracy o the CFD method in studying the thermal-hydraulic perormance o the PCHE is proved. 32.5 Table 3. Relative Deviation or Heat Transer and Pressure Drop at Fin Angle 32.5 Ishizuka Ishizuka Ngo Ngo h or Nu +21.1% +12.9% +33.% +2.6% -21.1% -13.8% +8.3% -1.9% -45.7% -55.8% -53.1% -52.3% 4. Table 4. Relative Deviation or Heat Transer and Pressure Drop at Fin Angle 4. Ishizuka Ishizuka Ngo Ngo h or Nu +38.5% +37.7% +56.6% +58.6% -19.7% -2.4% +5.8% -1.9% -1.4% -1.7% -35.2% -8.3% 4 PCHE Channel Structure Inluence on Heat Transer and Pressure Drop In general, an eective way to improve the perormance o the heat exchanger is to increase the average heat transer temperature dierence and the low rate. However, some parameters such as temperature, pressure and low rate o and luid are oten given and cannot be changed in the engineering, thereore, changing the heat exchanger structure is the most reasonable way to increase the heat transer. For the zigzag PCHE, the channel width, in angle, and in length determine its basic structure, all o which have a certain eect on heat transer and pressure drop. 4.1 Channel Width Inluence on PCHE Heat Transer and Pressure Drop In order to explore the channel width inluence on heat transer and pressure drop o PCHE, this chapter is based on the 32.5 in angle and 4.5mm in length. Considering the conventional PCHE channel width is generally 1.-6.mm, 1.mm to 6.mm channel width was created with an increment o 1.mm. The mass low rate was set rom 3kg/h to 6kg/h with an increment o 3kg/h. The pressure drop is greater under large low conditions and it is easier to judge whether the pressure drop is beyond the design range. In general, the pressure drop is within 1.5% o the inlet pressure that meets the design requirements. As shown in Fig.1, the convective heat transer coeicient increases with the increase o mass low rate under the same channel width condition, showing a linear growth relationship. At the same mass low rate, the convection heat transer coeicient decreases signiicantly with the increase o channel width. In the range o 3kg/h to 6kg/h, 1mm channel width PCHE has a convective heat transer coeicient o 19-32 W/(m 2 K), which is an excellent heat transer perormance. The heat transer eiciency 9
drops sharply when the channel width is increased rom 1mm to 2mm. For the 6mm channel width, the heat transer coeicient varies rom 9-17 W/(m 2 K), ar less than the heat transer perormance o 1mm channel width, which is only about 5% o the ormer. -2 K -1 h/w m 35 3 25 2 15 1 5 1mm 2mm 3mm 4mm 5mm 6mm 3 35 4 45 5 55 6 G/kg h -1-2 K -1 h/w m 18 16 14 12 1 8 6 4 2 3 35 4 45 5 55 6 Figure.1. Channel Width Inluence on PCHE Heat Transer Three representative channel width are chosen or analysis due to the similar heat transer and low characteristics, as illustrated in Fig.11. It indicates that the convection heat transer coeicient is obviously larger near the channel corner. The luid low direction changes and the disturbance increases signiicantly at the corner. The heat transer was enhanced with the boundary layer destroyed. In addition, the low velocity with small channel width is larger at the same mass low rate, and the turbulent state is more likely to develop rom laminar to turbulent low, which also improves the convection heat transer. 1mm 2mm 3mm 4mm 5mm 6mm G/kg h -1 (a) 1mm, 3kg/h, side (b) 2mm, 3kg/h, side (c) 3mm, 3kg/h, side Figure.11. Convection Heat Transer Coeicient Distribution o the Typical Channel Width As illustrated in Figure.12, the PCHE pressure drop changes linearly with the increasing o mass low rate. When the mass low rate increases to a certain value, the pressure drop begins to change exponentially with the mass low rate increasing. Obviously, under the same low conditions, pressure drop o 1mm channel width is much higher than other channel width. Although PCHE with 1mm channel width achieves excellent heat transer, while there is a substantial increase in pressure drop. The pressure drop o 1mm channel width has reached about.4mpa at 3kg/h mass low rate, accounting or 3.5% o the inlet pressure. At the low rate o 6kg/h, the pressure drop reaches about 1.6MPa, which is 14.5% o the inlet pressure and beyond the design requirements. It is achieved with a channel length o 54 mm, while the actual channel length is 896 mm. Excessive pressure drop can lead to excessive consumption o the pump power, aecting the thermal cycle economy. For the PCHE o 2mm channel width, the pressure drop is 2-8kPa in the range o 3kg/h to 6kg/h, accounting or.6% -2.5% o the inlet pressure on the side, which shows that 2mm channel width is 1
within the pressure drop design requirements. 2mm channel width PCHE can combine both heat transer and pressure drop, which is why a large variety o PCHE channel width are designed around 2mm. P/kPa 16 14 12 1 8 6 4 2 1mm 2mm 3mm 4mm 5mm 6mm 3 35 4 45 5 55 6 G/kg h -1 3 35 4 45 5 55 6 Figure.12. Channel Width Inluence on PCHE Pressure Drop For 3mm-6mm channel width PCHE, the pressure drop is small, but the heat transer perormance is poor. It indicates rom Fig.13 that the luid with the 2mm channel width starts to have some reversed lows at the corner, and at the 3mm channel width, the luid shows clear swirl lows, reversed lows and eddies at the corner, which causes a certain extent o pressure drop, thus deteriorating the heat transer perormance. P/kPa 2 18 16 14 12 1 8 6 4 2 1mm 2mm 3mm 4mm 5mm 6mm G/kg h -1 However, it can be ound that the luid velocity with 1mm channel width is signiicantly higher than that o 2mm and 3mm. In combination with Figure.12, the pressure drop caused by wall resistance at high velocity is signiicantly more than that caused by the reversed lows at low velocity. The high mass low rate is the reason that causes the pressure drop o 1mm channel width larger than other channel width, but also makes the heat transer perormance o 1mm channel width better. (a) 1mm, 3kg/h, side (b) 2mm, 3kg/h, side (c) 3mm, 3kg/h, side Figure.13. Velocity Vector Distribution o the Typical Channel Width Ater the above analysis, or some conditions that may enhance the pressure drop, such as large low conditions and long channel length, large channel width should be used. When the mass low rate is relatively small, or the channel length is relatively short, the pressure drop is more easily to meet the design requirements, small channel width should be considered in order to improve the heat transer perormance 4.2 Fin Angle Inluence on PCHE Heat Transer and Pressure Drop In order to explore the inluence o in angle on heat transer and pressure drop o PCHE, six low conditions vary rom 5kg/h-3kg/h were set based on PCHE o 2mm channel width, 4.5mm in length. The in angle increases rom 5 to 6 with an increment o 5. 11
As illustrated in Fig.14, under the same mass low rate, the large in angle causes more disturbance when the luid lows through the channel corner, which improves the heat transer perormance. Thereore, the heat transer perormance increases with the in angle gradually. It indicates that the curves o in angle and convection heat transer coeicient are approximately linear in the range o 5-2, 2-45 and 45-6. In the range o 5-2 and 45-6, the heat transer perormance is relatively lat with the in angle increasing. However, in the range o 2-45, the slope o the curve corresponding to the in angle and heat transer coeicient is relatively steep, and the heat transer increment is more obvious with the in angle increasing. It can be ound that the convective heat transer coeicient at 6 is close to twice o that at 5. Large in angle has more disturbance on the luid due to the large change o the low direction, so the convection heat transer coeicient is higher, as shown in Fig.15. -2 K -1 h/w m 9 8 7 6 5 4 3 2 1 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h 1 2 3 4 5 6 α/ -2 K -1 h/w m 45 4 35 3 25 2 15 1 5 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h 1 2 3 4 5 6 Figure.14. Fin Angle Inluence on PCHE Heat Transer α/ (a) 5, 3kg/h, side (b) 3, 3kg/h, side (c) 45, 3kg/h, side Figure.15. Convection Heat Transer Coeicient Distribution o the Typical Fin Angle The existence o the in angle enhances the luid disturbance signiicantly and improves the heat transer, but also causes great energy loss inevitably when the luid lows through the corner, that is, the pressure drop is increased. As shown in Fig.16, under the same mass low rate, the pressure drop also increases with the increase o in angle, while they show an exponential trend, that is dierent rom the tendency o heat transer coeicient and in angle. For a mass low rate o 3kg/h and 6 in angle, the corresponding pressure drop is 8.1kPa, accounting or.76% o the inlet pressure, which is within the pressure drop design requirements. However, as the mass low rate continues to increase or the channel length increases, it is possible to exceed the design requirement. 12
8 6 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h 7 6 5 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h P/kPa 4 P/kPa 4 3 2 2 1 1 2 3 4 5 6 α/ 1 2 3 4 5 6 Figure.16. Fin Angle Inluence on PCHE Pressure Drop As shown in Fig.17, it can be ound that the luid velocity o 5 in angle is the smallest, the 3 in angle is the second, and the 45 in angle is the maximum. Compared with the in angles o 5 and 3, the in angle o 45 has greater reversed low loss at the corners, resulted in a larger pressure drop. Combined with the Fig.14, the heat transer perormance changes little rom 45 because o the reversed low zone, while it is separated rom the main low zone and ormed a relatively "vacuum" state, thus leads to the poor heat transer perormance. α/ (a) 5, 3kg/h, side (b) 3, 3kg/h, side (c) 45, 3kg/h, side Figure.17. Velocity Vector Distribution o the Typical Fin Angle From the above analysis, i the pressure drop meets the requirements, large in angle has better heat transer perormance. In the range o 2-45 in angle, the heat transer perormance increases more obvious, while the pressure drop is relatively small, which can be ocused on when designing PCHE. 4.3 Fin Length Inluence on PCHE Heat Transer and Pressure Drop In order to explore the in length inluence on heat transer and pressure drop o PCHE, six low conditions vary rom 5kg/h to 3kg/h were set based on 2mm channel width and 3 in angle o PCHE. The 54mm channel length were separated into six categories, which are 27mm 2, 18mm 3, 9mm 6, 6mm 9, 3mm 18 and 2mm 27, which is illustrated in Fig.18. 13
(a) 27mm 2, 3kg/h, side (b) 18mm 3, 3kg/h, side (c) 9mm 6, 3kg/h, side (d) 6mm 9, 3kg/h, side (e) 3mm 18, 3kg/h, side () 2mm 27, 3kg/h, side Figure.18. Convection Heat Transer Coeicient Distribution o the Typical Fin Length As illustrated in Fig.19, with the decrease o the in length, the convective heat transer coeicient increases and reaches a peak value when the in length is 3mm. Thereater, the convection heat transer coeicient shows a decreasing trend as the in length continues to decrease. When the in length is relatively small, there are more disturbance when the luid lows through the corners, and the heat exchange is more suicient. When the in length continues to reduce, the corners becomes excessive and not sharp enough, the channels are more like the straight channels, which cannot match the heat transer perormance with zigzag channels. -2 K -1 h/w m 9 8 7 6 5 4 3 5kg/h 1kg/h 15kg/h 2kg/h 15kg/h 3kg/h -2 K -1 h/w m 45 4 35 3 25 2 15 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h 2 1 1 5 27 2 18 3 9 6 6 9 3 18 2 27 s/mm 27 2 18 3 9 6 6 9 3 18 2 27 Figure.19. Fin Length Inluence on PCHE Heat Transer As shown in Figure.2, at 3mm in length, the pressure drop reaches to the maximum, then is begins to decrease as the in length decreased. The mechanism is similar with that o heat transer. When the in length is small and the number o corners become ininite, while the low velocity and direction change little when low through the relatively straight channels, so the pressure drop is reduced and the heat transer perormance is also deteriorated. s/mm 14
P/kPa 7 6 5 4 3 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h P/kPa 5 4 3 2 5kg/h 1kg/h 15kg/h 2kg/h 25kg/h 3kg/h 2 1 1 27 2 18 3 9 6 6 9 3 18 2 27 s/mm 27 2 18 3 9 6 6 9 3 18 2 27 Figure.2. Fin Length Inluence on PCHE Pressure Drop It can be ound rom Fig.21 that the change o in length has little eect on the speed o low. The maximum velocity o these three types o in length is about 3m/s and about 15-2m/s on average. However, the 3mm 18 reversed low area is larger than 2mm 27, and the corresponding pressure drop is also larger. At the same time, it can be ound that the mainstream area o 3mm 18 is longer than 2mm 27. While the mainstream area o 2mm 27 is similar to the straight low, thus the heat transer perormance is not as good as 3mm 18. The ideal in length should change the luid low direction as much as possible. s/mm (a) 18mm 3, 3kg/h, side (b) 3mm 18, 3kg/h, side (c) 2mm 27, 3kg/h, side Figure.21. Velocity Vector Distribution o the Typical Fin Length Based on the above analysis, it indicates that the small in length has better heat transer perormance, meanwhile the increase o pressure drop is not large. But the in length is not as small as possible, there is also a peak perormance, around 3mm in length has the best overall perormance. 5 Conclusion This paper is based on the PCHE model o TIT. CFD method was used and compared with the experiments, which proves it accuracy and rationality. To maximize the heat transer perormance and minimize the pressure drop, channel structure inluence on PCHE was studied, which are summarized as ollows: The increase o channel width decreases the heat transer and the pressure drop simultaneously. Under the pressure drop design requirements, small channel width has the best heat transer eectiveness and 1.-2.mm channel width is recommended. The increase o in angle increases the heat transer and the pressure drop simultaneously. The heat transer increases little in 5-2 and 45-6 in angle, and the pressure drop grows ast in 45-6 15
in angle. The 2-45 in angle is avorable or it combines the good heat transer eectiveness and small pressure drop. The heat transer and pressure drop increases with the decrease o in length to a certain value, and then the trend becomes opposite. 3mm is the turning point in length and is recommended or its optimal heat transer perormance and limited pressure drop. 6 Reerences 1. Baek S, Kim JH, Jeong S, et al. Development o highly eective cryogenic printed circuit heat exchanger (PCHE) with low axial conduction[j]. Cryogenics, 212, 52(7): 366-374. 2. Gezelius K. Design o compact intermedia heat exchangers or gas cooled ast reactors [D]. Massachusettes: Massachusettes institute o technology, 24. 3. Southall, D., Le Pierres, R., Dewson, S. J., Design Considerations or Compact Heat Exchangers, Proceedings o ICAPP 8, Anaheim, CA., June 8-12, 28, Paper 89 4. Nikitin, K, Kato Y, Ngo TL. Printed circuit heat exchanger thermal hydraulic perormance in supercritical CO2 experimental loop [J]. Int. J. Rerig, 26, 29: 87 814. 5. Ngo T L, Kato Y, Nikitin K, et al. Heat transer and pressure drop correlations o microchannel heat exchangers with S-shaped and zigzag ins or carbon dioxide cycles [J]. Experimental Thermal & Fluid Science, 27, 32(2): 56-57. 6. Kim DE, Kim MH, Cha JE, et al. Numerical investigation on thermal-hydraulic perormance o new printed circuit heat exchanger model [J]. Nuclear Engineering and Design, 28, 238(12): 3269-3276. 7. Tsuzuki N, Kato Y, Ishiduka T. High perormance printed circuit heat exchanger [J]. Applied Thermal Engineering, 27, 27(1): 172-177. 8. Ishizuka T, Kato Y, Muto Y. Thermal-hydraulic characteristics o a printed circuit heat exchanger in a supercritical CO2 loop [C]. Avignon, France: The 11th International Topical Meeting on Nuclear Reactor Thermal Hydraulics, 25, October 2 6. 9. Kim S G, Lee Y, Ahn Y, et al. CFD aided approach to design printed circuit heat exchangers or supercritical CO2, Brayton cycle Application [J]. Annals o Nuclear Energy, 216, 92: 175-185. 1. DE Kim, MH Kim. Numerical investigation on thermal hydraulic perormance o new printed circuit heat exchanger model [J]. Nuclear Engineering and Design, 28, 238: 3269 3276. 11. Kim I H, Sun X. CFD study and PCHE design or secondary heat exchangers with FLiNaK-Helium or SmAHTR [J]. Nuclear Engineering & Design, 214, 27(5): 325-333. 12. John E.. Compact Heat Exchangers Selection, Design and Operation [M]. Pergamon: 21. 16