Experimental study on heat transfer of jet impingement with a moving nozzle

Accepted Manuscript Experimental Study on Heat Transfer of Jet Impingement with a Moving Nozzle X. Ai, Z.G. Xu, C.Y. Zhao PII: DOI: Reference:

S1359-4311(16)32828-9 http://dx.doi.org/10.1016/j.applthermaleng.2017.01.004 ATE 9768

To appear in:

Applied Thermal Engineering

Received Date: Accepted Date:

27 October 2016 3 January 2017

Please cite this article as: X. Ai, Z.G. Xu, C.Y. Zhao, Experimental Study on Heat Transfer of Jet Impingement with a Moving Nozzle, Applied Thermal Engineering (2017), doi: http://dx.doi.org/10.1016/j.applthermaleng. 2017.01.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Experimental Study on Heat Transfer of Jet Impingement with a Moving Nozzle X. Ai, Z.G. Xu, C.Y. Zhao * Key Laboratory of Power Machinery and Engineering of Ministry of Education, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China *Corresponding Author, Email: [email protected] (C.Y. Zhao) Abstract Water jet impingement heat transfer is widely used in electronic component cooling, steelmaking, nuclear power plants and many other high heat transfer rate applications. This paper describes experiments with a jet from a moving nozzle impinging a surface using a stepping motor to control the nozzle. The effect of nozzle velocity on the heat transfer rates at different heat fluxes and flow rates are investigated. The experimental results show that a moving nozzle performs better than a fixed nozzle for reducing the maximum temperature difference of the heating surface and the average liquid film thickness, which results in steadier heat transfer rates and a more uniform temperature. Furthermore, a moving nozzle enhances the heat transfer in the convection by more than forty percent. Here, a higher nozzle velocity better enhances the heat transfer and temperature uniformity. Key words Enhancement, Heat transfer, Jet impingement, Moving nozzle, Uniform temperature Introduction Current trends in electronic component application show an increasing need for efficient heat removal in microminiaturization systems with this need predicted to increase into the foreseeable future [1, 2]. For heat dissipation at high heat flux conditions, jet impingement is one of the most effective methods to 1

enhance the local heat transfer coefficient, which results in better cooling. Therefore, it is important to understand the heat transfer for jet impingement processes. Jet impingement heat transfer has been the subject of numerous studies over the past decade. There are various parameters that influence the heat transfer in the jet impingement processes, including the heat flux, flow rate, inlet pressure, nozzle size and working medium properties. In free-surface jet impingement on a planar surface, the heat transfer regions can be divided into five parts, with each part having its own empirical formula for determining the Nusselt number [3]. Furthermore, based on the empirical formula, the Reynolds number, Re, and Prandtl number, Pr, emerge as the most important parameters affecting the heat transfer. Merci et al. [4-6] researched the effects on the heat transfer of the Reynolds number and the distance between the nozzle and the surface. Here, the Reynolds number significantly influenced the heat transfer. Under identical distance conditions, a higher Reynolds resulted in a higher heat transfer coefficient. Merci also claimed that for the same Reynolds number, the heat transfer could be affected by the distance between the nozzle and the surface, with the best heat transfer when the distance was twice the diameter of the nozzle. In 2001, Hwang et al. [7] experimentally investigated the flow and heat transfer characteristics of an impinging jet that was controlled by vortex pairing in pulsed jets. They showed that flow excitation could either hinder or enhance the heat transfer. In a subsequent experimental study in 2003, Hwang and Cho [8] showed that both the frequency and excitation levels of acoustically excited impinging jets were important in heat transfer enhancement. Öztekin [9,10] suggested that surface roughness could also affect the heat transfer coefficient. An increase in the surface roughness enhanced the heat transfer. This conclusion was in agreement with the results of Dou et al. [11,12], who studied the heat transfer of jet impingement using steel plates with different thicknesses and surface roughnesses. By utilizing the PIV technique, Yang et al. [13,14] investigated the jet impingement

heat transfer, including the flow of the working medium, the 2

velocities of different points on the surface and the influence of different surface shapes, such as flat and curved surfaces. Himadri et al. and Senter et al. [15-17] both studied jet impingement heat transfer on moving surfaces, claiming that a moving surface influenced the flow characteristics, heat transfer, hydraulic jump, temperature distribution and heat transfer coefficient. What differs between our works and Himadri’s is that the size of their working surface is much bigger than ours and their controlled sensitivity as well as accuracy is much lower than ours. Besides, their works mainly focused on the hydrodynamic characteristic such as the hydraulic jump, and there are few detailed data of heat transfer process in their experiments. However, our works mainly concentrated on the heat transfer performance rather than the hydrodynamic characteristic. The majority of existing jet impingement studies are based on fixed nozzle experiments and simulations, showing that the substrate temperature distribution is inverted bell shape. This means that the temperature is lower in the middle and higher on both sides, which lead to non-uniform temperature distribution of the substrate. However, a non-uniform thermal load may influence the material properties. As such, improving the temperature distribution uniformity on the surface has important research significance. Also, the scour capability of the liquid film on heat transfer surfaces can be enhanced by increasing the Reynolds number [18-20]. In the current experiments, a fixed nozzle is replaced with a moving nozzle, which improves the scour capability, thereby enhancing the heat transfer. With this setup, the effects of nozzle velocity on the heat transfer for the jet impingement process at different heat fluxes and flow rates are determined. The effects of different kinds of substrates and nozzles on the heat transfer are also investigated and compared. 1. Experimental apparatus Figure 1 shows the experimental apparatus, which included a test chamber, main heater, testing samples, water jet nozzle, stepping motor, pipeline system and measurement system.

3

The stainless steel chamber (100 mm diameter x 100 mm height) contained the experimental equipment and liquid. Each side of the chamber had a quartz glass which served as windows to provide a light source and also to enable observation. A rectangular hole 10 mm deep × 30 mm in diameter was drilled into the center of the bottom chamber for the main heater. Teflon and epoxy glue filled the gap between the bottom chamber and the heater for sealing and insulation. The nozzle moved on the slide way when the stepping motor operated (Fig. 1b). The stepping motor rotation speed that determined the nozzle velocity was controlled by a Programmable Logic Controller (PLC)which is the functional module to control mechanical components as well as distance of moving. The piping included the tubes, pump, pressure sensor, valve and flow meter. The maximum pumping pressure was 1.1 MPa, with a flow meter that measured from 0 to 0.6 L/min. The main heater included two segments (Fig. 2a) [21]. The upper part was 10 x 30 x 4 mm (length x width x thickness). Two electric heating wires were brazed into the lower part of the heater. A voltage transformer supplied the power for the heater. The power was adjusted and measured by a power meter. The tested copper substrate was welded using tin onto the upper surface of the heater to eliminate the thermal contact resistance between the heater and sample. A total of eight T-type thermocouples were used in the experiment. Figure 2b, 2c and 2d shows the temperature measurement systems of the main heater and the copper substrates. Three T-type thermocouples (T1, T2 and T3) were affixed onto the side of the copper block to calculate the heat flux from the main heater. The remaining five T-type thermocouples (T4, T5, T6, T7 and T8) were welded onto the center of the copper samples to measure the mean wall temperatures of the substrate. The thermocouple signals were recorded by an NI data acquisition system. 2. Data reduction and uncertainty analysis The copper substrate was welded onto the top surface of the heater by a thin tin film to reduce the contact resistance. Grooves could not be used to mount the 4

thermocouples to the center of the heater top surface because the contact thermal resistance would increase and the uniformity of the heater surface temperature field would deteriorate. A one-dimensional temperature distribution in the upper part of the heater (along the vertical direction) was assumed because the side surfaces of the main integrated copper heater were thermally insulated by glass wool and the thermal conductivity of copper is high. Hence, an approximately linear temperature distribution was assumed based on the measured temperatures T1, T2 and T3. The heat flux on the upper surface of the heater was obtained by extrapolation as: q  ks

dT dz

(1) w

where ks is the copper thermal conductivity;

dT dz

is the temperature gradient on w

the heater upper side; and z is the coordinate perpendicular to the substrate surface. It was difficult to weld thermocouples onto the surface of the copper substrate for the present study; therefore, thermocouples were affixed inside the center of the substrate. Equation (2) shows the average value of the T4, T5,T6, T7 and T8, which was taken as the mean temperature of the substrate due to the high thermal conductivity of copper. (2)

Tw = (T4 +T5 +T6 +T7 +T8 ) 5

Thus, the heat transfer coefficient was defined based on the thermal conduction inside the material as:

h = q (Tw -Ts )

(3)

where Ts is the saturation temperature of water at atmospheric pressure. The surface temperature, surface superheat and pool liquid temperature had maximum uncertainties of 0.1°C for T-type thermocouples. The relative uncertainty of h was calculated using [22]: 2

2

 q   Tw   Ts  h        h  q   Tw  Ts   Tw  Ts 

2

(4)

The thermocouple locations gave the uncertainty of q as: 5

q (2 z1  z2  z3 )T1  (2 z2  z1  z3 )T2  (2 z3  z1  z2 )T3  q (2 z1  z2  z3 )T1  (2 z2  z1  z3 )T2  (2 z3  z1  z2 )T3

(5)

where z1, z2 and z3 are the distances of the three thermocouples T1, T2 and T3, respectively, from the upper surface of the main heater. The maximum relative uncertainty of h was 8 % from a standard uncertainty analysis. Before each experimental study on the different substrates, tin was placed on the upper surface of the main heater. Firstly, the main heater was switched on to increase the temperature of the upper surface to the melting point of tin. Next, the copper sample was adhered to the tin. Finally, the heater was switched off and the tin solidified by natural convection cooling. The tin thickness between the heater surface and the foam substrate was approximately 0.05 mm, which was less than 2.5 % of the metallic foam thickness. Thus, the metallic sample was tightly welded onto the heater surface. Furthermore, the contact thermal resistance was essentially eliminated. At the beginning of the experiment, the heater was switched on to start the heat flux (50 W/m2). After the copper substrate temperature reached 100°C, the pump was turned on and the water impinged onto the substrate through the nozzle without moving. Data were collected by the NI data acquisition system. The pump was turned off when the variation of the sample substrate temperature was within 0.3°C (approximately two minutes). After the copper substrate temperature reached 100°C, the pump and stepping motor were turned on until the temperature stabilized. The experiments were repeated at different nozzle velocities (0 to 20 mm/s), different heat fluxes (50 to 300 W/m2) and different flow rates (4 to 8 ml/s). The Reynolds number ranged from 5060 to 10120.

3. Results and discussions 3.1 Effects of nozzle velocity on the heat transfer at different heat fluxes The main experimental subject in this study is the effect of the finned substrate (Fig. 2c). The temperature distribution on the copper substrate on the heater is illustrated in Fig. 3 as a function of distance from the stagnation point for 6

a nozzle with a diameter of 1 mm and a Reynolds number of 7590. The heat flux is 150 W/cm2 in Fig. 3a and 300 W/cm2 in Fig. 3b. The moving velocity of the nozzle ranges from 0 to 20 mm/s. Figure 3 shows that when the nozzle is fixed, the temperature distribution is an inverted bell shape. However, as the nozzle moves, the temperature distribution of the copper substrate becomes uniform, with a 1°C variation. Furthermore, a higher nozzle velocity results in a more uniform temperature distribution and a lower average substrate temperature. Comparison of Fig. 3a with Fig. 3b shows that the effect of the nozzle velocity on the temperature distribution is more significant as the heat flux increases. The maximum substrate temperature difference at the same nozzle velocity increases from 3.9 to 5.5°C as the heat flux increases from 150 to 300 W/cm2. Figures 4 and 5 show that the average substrate temperature and Nusselt number increase with increasing nozzle velocity at different heat fluxes. The Nusselt number is defined as: Nu 

hl

(6)



Where h is the convective heat transfer coefficient, λ is the thermal conductivity of the fluid and l is the surface length. The Reynolds numbers are 7590 and 10120. As the nozzle velocity increases, the average temperature and Nusselt number both improve, i.e., the average temperature is lower and the Nusselt number is higher, regardless of the heat flux. Furthermore, compared with a fixed nozzle, the effects of the moving nozzle are more significant at higher heat flux. Figures 4 and 5 show that the maximum average temperature difference is approximately 7°C and that the Nusselt number increases 31 %, to almost 1250 at the same heat flux. Relative to a fixed nozzle, a moving nozzle significantly enhances the heat transfer because the nozzle movement improves the scour capability of the liquid film on the substrate. For jet impingement heat transfer, Eq. (7) shows that the forced-convection heat transfer is the most mechanism driving the heat transfer between the working medium and substrate. 7

Q

λ film l film

Nu film  Afilm (Tw  T film )

(7)

Due to the complexity of the jet impingement process, its flow field can be divided into five different regions as shown in Fig. 6 [3, 23]. In addition, each region has its own heat transfer process with Nufilm of each region given by: RegionⅠ: Nu film  0.745 Pr1 / 3 Re0.5

(8)

1/ 3 0.5 RegionⅡ： Nu film  0.668 Pr Re  r / d 

0.5

(9)

3 Region Ⅲ： Nu film  1.5874Pr 1/3Re1/3  25.735 r / d  Re 1  0.8566   

RegionⅣ： Nu film

 4  r 2 272 H   1  b     1   525d   Pr Re  d 

2/3

(10)

1

(11)

where d is the nozzle diameter; H is the distance between the nozzle and the substrate; r is the distance from the stagnation point to the test point; and Nufilm is Nu of the liquid film. These equations show that the Nusselt number is a function of the Reynolds number and the Prandtl number, with the working medium and nozzle size also closely related to the development of the jet flow. Equations (8) to (11) show that jet impingement heat transfer with a fixed nozzle has the best heat transfer in region I. As the working medium on the substrate develops, Nufilm decreases from region I outwards. However, a moving nozzle improves the scour capability of the liquid film, which adequately compensates for the heat transfer in the lower Nufilm regions. Thus, the entire substrate surface has more high heat transfer areas with higher Nufilm since regions I or II replace regions IV or V. Also, both the heat transfer and temperature distribution uniformity are significantly enhanced for the entire jet impingement process. However, the heat transfer rate (Q) of the jet impinging is inversely proportionate to the average thickness of the liquid film (lfilm) (Eq. (7)). The liquid film becomes thicker as the working medium collects from region I to V (Fig. 6). Here, the moving nozzle changes the development of the working medium film layer, reducing the average liquid film thickness, and enhancing the heat transfer 8

for the entire jet impingement process. Figure 7a shows that when the nozzle is fixed, there is a layer of liquid film on the surface. However, after the nozzle moves above the surface, the liquid film almost entirely disappears (Fig. 7b and 7c); thereby, confirming this explanation.

3.2 Effects of nozzle velocity on the heat transfer at different Reynolds numbers Figure 8 shows that the temperature distribution on the copper substrate on the heater (after the jet impingement process) stabilizes for a heat flux of 50 W/cm2. The Reynolds number varies from 5060 in Fig. 8a to 10120 in Fig. 8b, with the nozzle velocity ranging from 0 to 20 mm/s. These curves are similar to the curves in Fig. 3 showing the same phenomenon, i.e., a higher nozzle velocity leads to a more uniform temperature distribution as well as a lower average substrate temperature. Furthermore, when the flow rate decreases, the effect of the moving nozzle is more prominent. Figure 8a and 8b shows that the maximum temperature difference of the substrate increases from 2 to 4°C as the Reynolds number decreases from 10120 to 7590 for the same nozzle velocity. As the nozzle velocity increases, the average substrate temperature and average Nusselt number vary at different flow rates as shown in Figs. 9 and 10. As the nozzle velocity increases, the heat transfer improves. At a Reynolds number of 10120, the maximum average Nusselt number difference is approximately 1050, which is an improvement of 28 %. However, when the Reynolds number decreases to 5060, the average Nusselt number increases almost 43%. This strongly suggests that a moving nozzle is more effective at lower Reynolds numbers. The influence of a moving nozzle on the heat transfer is more obvious at a lower Reynolds number because a decreasing Reynolds number means a decreasing outlet nozzle velocity. Thus, the scour capability is weaker on the substrate surface. Here, the moving nozzle significantly improves the scouring ability, thereby enhancing the heat transfer . However, when the Reynolds number 9

is sufficiently high, a fixed nozzle also has a good scour capability for the same conditions. Hence, there are only moderate enhancements to the heat transfer at higher Reynolds numbers.

3.3 Effects of nozzle velocity on the heat transfer with different substrates To compare the heat transfer between different kinds of substrates, a flat substrate was also experimentally studied as shown in Fig. 2d. Figure 11a shows the temperature distributions on the two substrates at different nozzle velocities for a heat flux of 150 W/cm2 and a flow rate of 6 ml/s. The finned substrate temperature is lower than the flat substrate temperature. This phenomenon is similar to previous research which showed that the fins improve the conduction and enlarge the heat transfer area (A) between the substrate and the working medium for the heat transfer. Also, when the nozzle is fixed, the temperature distribution on both substrates is an inverted bell shape with the finned substrate having a lower minimum temperature than the flat substrate. The maximum temperature difference from the stagnation point to both sides is 4.7°C for the finned substrate but only 2.1°C for the flat substrate. This is because the fins partially resist the working medium convection on the surface. Therefore, the scour capability of the finned substrate is worse than on the flat substrate if the nozzle is fixed. The moving nozzle enhances the heat transfer and makes the temperature distribution more uniform and steady. Figure 11b and 11c shows the variations of the average substrate temperature and average Nusselt number with increasing nozzle velocity. According to the experimental data, the finned substrate enhances the heat transfer better than the flat substrate with a lower average temperature and higher average Nusselt number because of the positive effects on the heat transfer of the fins. Furthermore, a comparison of the curves in Fig. 11b and 11c shows that the gradient on the finned substrate curve is bigger than that on the flat substrate. On the finned substrate, the maximum average Nusselt number increases by 10

approximately 650 from zero flow rate up to the maximum velocity, which is an increase of 20%. However, the average Nusselt number only increases 13% on the flat substrate for the same velocities. This means that an increase in the nozzle velocity enhances the positive effect of the fin. This tendency is due to the moving nozzle, which improves the scour capability and removes the negative influence of the fins on the convection. Hence, the experiments show that the moving nozzle enhances the heat transfer more on a finned substrate than on a flat substrate.

3.4 Effects of nozzle velocity on the heat transfer for different nozzle diameters The effects of different nozzle diameters (1 to 0.8 mm) were also experimentally studied. Figure 12a and 12b shows the variations of the average substrate temperature and the average Nusselt number for different nozzle diameters with a heat flux of 150 W/cm2 and a flow rate of 6 ml/s. The same regular pattern is observed, with the moving nozzle increasing the heat transfer irrespective of the nozzle diameter. However, for the same conditions, a smaller nozzle has better heat transfer. This phenomenon occurs because a smaller nozzle has a higher Reynolds number, which most strongly affects the scour capability and Nufilm (Eqs. (7) to (11)). Hence, a smaller nozzle has better heat transfer than a larger diameter nozzle. 4. Conclusions The heat transfer characteristics of a water jet impinging a surface with a moving nozzle were experimentally investigated. The effects of nozzle velocity on the heat transfer at different heat fluxes and Reynolds numbers is as follows: (1) A moving nozzle enhances the heat transfer uniformity, which leads to a more uniform temperature distribution, with increased uniformity as the nozzle velocity increases. (2) A moving nozzle enhances the heat transfer by reducing the average surface temperature and the liquid film thickness. A higher nozzle velocity

11

(more than 10 mm/s) results in a forty percent increase in the average Nusselt number. (3) The experiments showed that the heat flux increases from 50 to 300 W/cm2 as the Reynolds number increases from 5060 to 10120. The moving nozzle more effectively enhances the heat transfer at higher heat fluxes (over 200 W/cm2) and at lower Reynolds numbers (below 7590). (4) The addition of fins to the substrate enhances the heat transfer with the moving nozzle giving better heat transfer on a finned substrate than on a flat substrate. Also, for the same conditions, decreasing the nozzle diameter enhances the heat transfer.

Acknowledgments This work was supported by the National Key Basic Research Program of China (973 Project: 2013CB228303) and the National Natural Science Foundation of China (Grant No. 51576126) and the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 51521004). Nomenclature Afilm

- Heat transfer area, m2

d

- Nozzle diameter, mm

h

- Heat transfer coefficient, W·cm-2·K-1

Ks

- Heat conductivity coefficient of copper, W·m-1·K-1

Lfilm

- Liquid film thickness, m

m

- Flow rate, ml·s-1

q

- Heat flux, W·cm-2

Q

- Heat transfer rate, W

r

- Distance between middle to any point, mm

R

- Distance between middle to edge, mm

Ti,i=1,2…

- Thermal couple temperature, °C 12

Tw

- Surface temperature, °C

Ts

- Working medium temperature, °C

v

- Moving nozzle velocity, mm·s-1

Zi,i=1,2,3

- Distance between two thermal couples, mm

Nu

- Nusselt number

Re

- Reynolds number

References [1] Moore, G. E., Progress in digital integrated electronics [C]. IEEE International Electron Devices Meeting, 1975: 11-13. [2] Moore, Gordon E. Cramming more components onto integrated circuits, Reprinted from Electronics, volume 38, number 8, April 19, 1965, pp. 114 ff.IEEE Solid-State Circuits Newsletter 3.20 (2006): 33-35 [3] Ma, C. F., Zhao, Y. H., Masuoka, T., Gomi, T. Analytical study on impingement heat transfer with single-phase free-surface circular liquid jets. Journal of Thermal Science 5.4 (1996): 271-277. [4] Merci, B., Heat transfer predictions with a cubic k-ε model for axisymmetric turbulent jets impinging onto a flat plate. International Journal of Heat and Mass Transfer. 46 (2003), 469-480. [5] Merci, B., De Langhe, C., Vierendeels, J. and Dick, E.. A quasi-realizable cubic low-Reynolds Eddy-viscosity turbulence model with a new dissipation rate equation. Flow Turbulence and Combust. 66 (2) (2001), 133-157. [6] Merci B. and Dick, E..Predictive capabilities of an improved cubic k–e model for inert steady flows. Flow Turbul. Combust. 68 (4) (2002), 335-358. [7] Hwang, S. D., C. H. Lee, and H. H. Cho. Heat transfer and flow structures in axisymmetric impinging jet controlled by vortex pairing. International Journal of Heat and Fluid Flow 22.3 (2001): 293-300. [8] Hwang, S. D., and H. H. Cho. Effects of acoustic excitation positions on heat transfer and flow in axisymmetric impinging jet: main jet excitation and shear

13

layer excitation. International Journal of Heat and Fluid Flow 24.2 (2003): 199209. [9] Öztekin E, Aydin O, Avcı M. Heat transfer in a turbulent slot jet flow impinging on concave surfaces. International Communications in Heat and Mass Transfer. 44 (2013), 77-82. [10] Kondo, Y., Behnia, M., Nakayama, W., Matsushima, H. Optimization of finned heat sinks for impingement cooling of electronic packages. Journal of Electronic Packaging 120.3 (1998): 259-266. [11] Dou, R., Wen, Z., Zhou, G., Liu, X., Feng, X. Experimental study on heattransfer characteristics of circular water jet impinging on high-temperature stainless steel plate. Applied Thermal Engineering 62.2 (2014): 738-746. [12] Dou R, Wen Z, Zhou G. 2D axisymmetric transient inverse heat conduction analysis of air jet impinging on stainless steel plate with finite thickness[J]. Applied Thermal Engineering 93 (2016): 468-475. [13] Xu Y, Feng L H, Wang J J. Experimental investigation of a synthetic jet impinging on a fixed wall. Experiments in fluids 54.5 (2013): 1-13. [14] Ma L Q, Feng L H. Experimental investigation on control of vortex shedding mode of a circular cylinder using synthetic jets placed at stagnation points." Science China Technological Sciences 56.1 (2013): 158-170. [15] Senter, J., and C. Solliec. Flow field analysis of a turbulent slot air jet impinging on a moving flat surface. International Journal of Heat and Fluid Flow 28.4 (2007): 708-719. [16] H. Chattopadhyay, S. K. Saha. Turbulent flow and heat transfer from a slot jet impinging on a moving plate. International Journal of Heat and Fluid Flow 24 (2003) 685–697 [17] Gradeck, M., Kouachi, A., Dani, A., Arnoult, D., Borean, J. L. Experimental and numerical study of the hydraulic jump of an impinging jet on a moving surface. Experimental Thermal and Fluid Science 30 (2006) 193–201 [18] Reddy A V. Experimental and numerical simulation study of heat sinks with impingement flow at high Reynolds numbers[C]//Semiconductor Thermal 14

Measurement and Management Symposium, 2003. Ninteenth Annual IEEE. IEEE, 2003: 176-178. [19] K. Jambunathan, E. Lai, M.A. Moss, B.L. Button. A review of heat transfer data for single circular jet impingement. International Journal of Heat and Fluid Flow 13.2 (1992): 106-115. [20] D.W. Colucci, R. Viskanta. Effect of nozzle geometry on local convective heat transfer to a confined impinging air jet. Experimental Thermal and Fluid Science 13.1 (1996): 71-80. [21] Z.G. Xu, Z.G. Qu, C.Y. Zhao, W.Q. Tao. Pool boiling heat transfer on opencelled metallic foam sintered surface under saturation condition. International Journal of Heat and Mass Transfer 54.17 (2011): 3856-3867. [22] Z.G. Xu, Z.G. Qu, C.Y. Zhao, W.Q. Tao. Experimental study of pool boiling heat transfer on metallic foam surface with U-shaped and V-shaped grooves. Journal of Enhanced Heat Transfer 19.6 (2012): 549. [23] Lienhard, John. Liquid jet impingement. Annual Review of Heat Transfer6.6 (1995).

15

(a) Schematic

(b) Experimental apparatus Figure 1. Experimental system

16

(a) Heater construction

(b) Thermocouple arrangement

(c) Finned substrate

(d) Flat substrate

Figure 2. Heater construction, thermocouple arrangement and substrates 17

(a) q =150 W·cm-2

(b) q =300 W·cm-2 Figure 3. Surface temperature distributions at different heat fluxes 18

(a) Re =7590

(b) Re =10120 Figure 4. Average surface temperatures at different heat fluxes

19

(a) Re =7590

(b) Re =10120 Figure 5. Average Nusselt numbers at different heat fluxes

20

Figure 6. Schematic diagram of the jet impingement

(a) Liquid film with a fixed nozzle

(b) Moving nozzle, t=0.5 s

(c) Moving nozzle, t=2.5 s

Figure 7. Liquid film variation

21

(a) Re =7590

(b) Re =10120 Figure 8. Surface temperature distributions at different Reynolds numbers

22

(a) q =50 W·cm-2

(b) q =200 W·cm-2 Figure 9. Average surface temperatures at different Reynolds numbers 23

(a) q =50 W·cm-2

(b) q =200 W·cm-2 Figure 10. Average Nusselt numbers at different Reynolds numbers 24

(a) Temperature distribution

(b) Average surface temperature

25

(c) Average Nusselt number Figure 11. Comparison of the heat transfer effects on finned and flat substrates

26

(a) Average surface temperature

(b) Average Nusselt number Figure 12. Heat transfer with different nozzle diameters

27

Figure captions Figure 1. Experimental system Figure 2. Heater construction, thermocouple arrangement and substrates Figure 3. Surface temperature distributions at different heat fluxes Figure 4. Average surface temperatures at different heat fluxes Figure 5. Average Nusselt numbers at different heat fluxes Figure 6. Schematic diagram of jet impinging Figure 7. Liquid film variation Figure 8. Surface temperature distributions at different Reynolds numbers Figure 9. Average surface temperatures at different Reynolds numbers Figure 10. Average Nusselt numbers at different Reynolds numbers Figure 11. Comparison of the heat transfer effects on finned and flat substrates Figure 12. Heat transfer with different nozzle diameters

28