A New Cooling Technique for Military Systems; Transpiration Cooling

Transkript

1 Savunma Bilimleri Dergisi The Journal of Defense Sciences Mayıs/May 2016, Cilt/Volume 15, Sayı/Issue 1, ISSN (Basılı) : ISSN (Online): A New Cooling Technique for Military Systems; Transpiration Cooling Mustafa KILIÇ 1 Abstract Development of new technologies causes an increase of heat loads of electronic systems in military systems. So applying new techniques to cool electronic systems is an obligation. For this reason, present study is focused on investigation of heat transfer from a porous plate by cooling of air and surface with transpiration cooling. Effects of Reynolds number of hot air (Re= 3035, 3200, 3300, 3580), temperature of air (T air = 47, 57, 67, 77 ºC) on local wall temperature and cooling efficiency of porous flat and the system inside a rectangular channel with air as a hot gas stream and water as a coolant were investigated numerically and experimentally. Numerical model was validated and formed with RNG k-ε model. It was observed that Reynolds number causes an increase on surface temperature and a decrease on cooling efficiency of porous plate. Increase of 15.2% on Reynolds number causes a decrease of efficiency of the system of 6.9%. Decreasing of inlet temperature causes not only a significant decrease on surface temperature but also a slight increase on cooling efficiency of the system. Increasing air inlet temperature of 9.4 % causes only an increase on efficiency of the system of 0.7%. It was also seen that numerical results generally have a good approximation with experimental results. Keywords: Heat Transfer, Transpiration Cooling, Porous Plate, CFD. 1 Address: Assistant Prof.Dr., Adana Science and Technology University, Faculty of Engineering and Natural Sciences, Mechanical Engineering Department, Adana, mkilic@adanabtu.edu.tr. Retrieved Date: Accepted Date:

2 202 Kılıç Askerî Sistemler İçin Yeni Bir Soğutma Tekniği; Terleme ile Soğutma Öz Gelişen yeni teknolojiler askerî sistemlerde kullanılan elektronik sistemlerin ısıl yüklerinin belirgin şekilde artmasına sebep olmaktadır. Böylece elektronik sistemlerin soğutulması için yeni teknikler kullanmak bir zorunluluk haline gelmektedir. Bu nedenle; bu çalışma havanın ve gözenekli (poros) plaka yüzeyinin terleme ile soğutma tekniği kullanılarak soğutulmasının incelenmesini hedeflemiştir. Dikdörtgen bir kanal içerisinde, havanın sıcak gaz akışı, suyun soğutucu olarak kullanıldığı bir durumda; havanın Reynolds sayısının (Re= 3035, 3200, 3300, 3580), hava giriş sıcaklığının (T hava = 47, 57, 67, 77 ºC) yerel plaka yüzey sıcaklığı ve gözenekli plaka ve sistem soğutma verimine etkileri sayısal ve deneysel olarak incelenmiştir. Doğrulanmış sayısal model olarak RNG k-ε türbülans modeli kullanılmıştır. Reynolds sayısı arttıkça, plaka yüzey sıcaklığında artış ve gözenekli plaka soğutma veriminde ise bir düşüş olduğu gözlenmiştir. Reynolds sayısında % 15,2 lik bir artış, sistem soğutma veriminde % 6,9 luk bir düşüşe sebep olmuştur. Giriş hava sıcaklığının azaltılması yalnızca yüzey sıcaklığında önemli bir düşüşe sebep olmamış, aynı zamanda sistem soğutma veriminde de kısmi bir artışa sebep olmuştur. Giriş hava sıcaklığının % 9,4 lük bir artış sistem soğutma veriminde yalnızca % 0,7 lik bir artışa sebep olmuştur. Sayısal sonuçların genel olarak deneysel sonuçlara oldukça iyi bir değerde yakınsadığı görülmüştür. Anahtar Kelimeler: Isı Aktarımı, Terleme ile Soğutma, Gözenekli Plaka, Hesaplamalı Akışkanlar Dinamiği. Introduction Transpiration cooling is considered as an attractive cooling technique. If passive and ablative protection are a way to ensure the thermal management, the active cooling, especially transpiration cooling, is probably the most efficient way to enable the structure withstanding of such large heat load. This method has been used to protect solid surface exposed to high-heat-flux, high-temperature environments such as hypersonic vehicle combustors, liquid rocket thrusters, gas turbine blades, water oxidation technology, the nose of aerospace vehicles during the atmospheric re-entry phase of their flight, shells, rockets and in many military applications. Transpiration cooling processes involve simultaneously two different heat transfer mechanisms: conduction through a solid plus

3 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), convection. By using transpiration cooling not only solid surface can be protected but also temperature of hot gas can be reduced. So it can also be used to cool air. In transpiration cooling process: fluid coolant is injected into porous matrix in the direction opposite to heat flux, at the same time absorbs the heat conducted into the solid matrix and transports heat flux through the convection passing pores, finally the coolant forms a thin film on the hot side surface to reduce the heat flux coming into the porous matrix and to cool the hot gas stream. Scope of transpiration cooling is shown in Fig.1. Figure 1. Scope of transpiration cooling Many experimental and numerical investigations on transpiration cooling were performed and can be found in the literature. Jiang, Yu, Sun and Wang (2004) investigated turbulent flow and heat transfer in a rectangular channel without and with transpiration cooling experimentally and numerically. Their results showed that the transpiration cooling greatly increases the boundary layer thickness and reduces the wall skin friction and increasing coolant blowing ratio sharply reduced both the wall temperature and the convection heat transfer coefficient. Liu, Jiang, Jin, and Sun (2010) investigated the transpiration cooling mechanisms for thermal protection of a nose cone experimentally and numerically for various cooling gasses. In their study: they used air, nitrogen, argon, carbon dioxide and helium as a hot gas stream. Their results showed that the injection rate strongly influenced the cooling effectiveness. The increase of the main stream inlet Reynolds number dramatically reduced the cooling effectiveness. And the coolant thermos physical properties, especially specific heat, most strongly influenced the cooling effectiveness. Liu, Jiang, Xiong, Wang (2013) investigated the flow and heat transfer characteristic of transpiration cooling through sintered porous flat plates with particle diameters d p = 40 and 90 μm experimentally and numerically with dry air as the coolant stream. They showed that the cooling effectiveness increased with increasing injection rate, the temperature distribution on the porous bronze plate was more uniform than that on the sintered stainless steel plates and the cooling

4 204 Kılıç performance for the porous wall with the smaller particle diameters was better. Arai and Suidzu (2012) investigated experimentally effects of the porous ceramic coating material such as permeability of cooling gas, thermal conductivity and adhesion strength. The mixture of 8 wt.% yttriastabilized zirconium and polyester powders was employed as the coating material, in order to deposit the porous ceramic coating onto Ni-based super alloy substrate in their study. They showed that porous ceramic coating has superior permeability for cooling gas and transpiration cooling system for gas turbine could be achieved by using porous ceramic coating. He, Wang, Xu, Wang (2013) investigated new conversation equation for mass, momentum and energy to describe the performances of fluid flow, heat absorption and phase change in porous matrix. Their model s main differences from previous models are firstly, considering the compressibility of vapor in the momentum and energy equations, secondly, adding a term of the momentum transfer caused by liquid phase change into the momentum equation of vapor and liquid phases in two-phase region, finally in the energy equation of two-phase region, taking the variations of temperature and pressure into account. Their results showed that with an increase in heat flux and decrease in coolant mass flow rate the temperature difference over the two-phase region falls and a higher external heat flux or lower coolant mass flow rate will accelerate the phase change process and increase he area of two-phase region and vapor region. Wang, Zhao, Wang, Ma and Lin (2014) investigated the effect of different mainstream temperature, Reynolds numbers, and coolant injection ratio on transpiration cooling of the wedge shape nose cone with an equal thickness porous wall using liquid water as coolant. They obtained that the average temperature over the transpiration area falls with an increase in the coolant injection ratio, whereas the average cooling effectiveness rise and There is an optimal injection rate, at which the coolant passing through pores with liquid state when the driving force for the coolant injection is the minimum and the cooling effectiveness is high. Tsai, Lin, Wang, Lin Y., Su, and Yang (2009) investigated the transient cooling process in a sudden-expansion channel with the injection of cold air from the porous bottom wall experimentally. They categorized flow features to four cooling patterns; the recirculation pattern, the elevated recirculation pattern, the transpiration pattern and film pattern. Langener, Wolserdorf, Selzer, and Hald (2012) investigated transpiration cooling applied to flat C/C material under subsonic main-flow conditions. In their study; main-stream Mach number ranged from Mg= and totals temperature was 523 K. Air, argon and helium were used as coolants. They showed that thickness of sample and main stream total

5 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), temperature did not affect the cooling efficiency. The coolant used and its specific capacity was the most influential parameter for the cooling efficiency. Zhao, Wang, Ma, Lin, Peng, Qu, and Chen (2014) presented a supersonic experimental investigation on transpiration cooling of a nose cone model with unequal thickness walled configuration using liquid water as coolant. Their experiments showed that ice cake appeared at the nose cone surface at main stream Mach number 2.0, stagnation pressure 174 kpa and mass flow rate kg/s with an enthalpy 300 kj/s. But ice cake disappeared when the stagnation pressure and enthalpy rise to 305 kpa and 1300 kj/kgm and the design of unequal thickness walled configuration nose cone is effective to solve the key issue of cooling stagnation point. So optimization of wall thickness should be based on mainstream condition, permeability and sample structure. Tsai, and Lee (2014a, 2014b) investigated the correlation between superheat levels and heat fluxes when used sintered powder structures as wicks. Their parameters were 45 μm, 75 μm, 150 μm of powder sizes and powder shapes of spherical, dendritic. Their results showed that smaller powders structures achieved higher effective thermal conductivities for both powder shapes. Spherical powder structures achieved twice the effective thermal conductivity of dendritic powder ones for each powder size. He, Guan, Gurgenci, Hooman, Lu, and Alkhedhair (2014) investigated performance of evaporative cooling with cellulose and Polyvinyl Chloride (PVC) corrugated media experimentally. The heat and mass transfer and pressure drop across the two media with three thicknesses (100, 200, 300 mm) were studied. Their results showed that the pressure drop range of the cellulose media is Pa while the pressure drops of the PVC media are much lower with the range of Pa, depending on the medium thickness, air velocity and water flow rate. The cooling efficiency of the cellulose media vary from 43% to 90% while the cooling efficiency of the PVC media are 8% to 65% depending on the medium thickness and air velocity. Polezhaev (1997) numerically investigated transpiration cooling for critical high-temperature turbine blade. He showed that transpiration gas-cooled blade concept had a potential to achieve the goal of 60% thermal efficiency for a gas turbine power plant. Trevino and Medina (1997) also obtained asymptotic and numerical results for the steady-state transpiration cooling of a thin porous flat plate in a laminar hot convective flow. The injection strength and its distribution along the plate are their parameters. They presented the asymptotic solution for the thermally thin wall regime and they showed that numerical results were in a good agreement with the asymptotic solution close to the asymptotic limits. Andoh and Lips (2003) presented an analytical approach

6 206 Kılıç method for determining the fluid flow and the heat transfer characteristics in the porous matrices. The governing parameters were the volumetric heat transfer coefficient, the equivalent thermal conductivity of the material and fluid flow characteristics. They obtained that an increase of 90% of the value of volumetric heat transfer coefficient involves a decrease from 45% to 54% of the downstream face temperature. And the best conductivity material assures a best evacuation of heat flux, assuming a best thermal protection of the wall. They tested their method by simulation and compared with those of literature. They obtained good agreement with the previsions. Cerri, Giovannelli, Battisti, and Fedrizzi (2007) compared transpiration cooling to effusion cooling. The diameter, density and spacing of the holes, adiabatic film efficiency and the coolant air consumption were parameters of their numerical study. They showed that transpiration-cooled components have proved an effective way to achieve high temperatures and erosion resistance for gas turbines. Effusion cooling on the other hand is relatively simple and more reliable technique offering a continuous coverage of cooling air over the component s hot surfaces. Liu, Xiong, Jiang, Wang, Sun (2013) investigated numerically the influence of the thermal conductivity and porosity of the porous wall and locally high heat fluxes on the hot surface and the location of a low porosity region within one and two layer porous walls on the temperature field. The local thermal nonequilibrium model was used to predict the temperature distribution in the solid. Their results showed that both the porosity and the thermal conductivity variations had a substantial effect on the temperature distribution. The reduction of the thermal conductivity and the decrease of the porosity reduced the influence of the outside boundary conditions to a thin layer. And the location of a low porosity obstructed region in the ceramic coating of the two layer model with the bronze base matrix had a pronounced effect on the temperature field. He and Wang (2014) studied on a simulation about one-dimensional and steady-state transpiration cooling problem with phase change numerically. Different heat flux and coolant injection were their parameters to analyze thickness of two-phase region and driving force of coolant injection. Their results showed that the thickness of the two-phase region does not vary with heat flux, when the plate is in vapor, two-phase and liquid phase, while the thickness is mainly influenced by coolant mass flow rate. And the pressure and capillary pressure at the bottom of the plate had the same trend so the capillary pressure was the main factor causing pressure change. Huang, Zhu, Xiong, and Jiang (2014) investigated cooling of a sintered porous struts using transpiration cooling in scramjets numerically. Methane was used as coolant fluid. Different strut

7 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), structure, material properties and coolant conditions were their parameters in their study. Main stream velocity was 2.5 Mach and temperature of it was 1920 K. Their results showed that increasing the coolant blowing ratio improves the cooling effectiveness. The strut surface cooling effectiveness first increased and then decreased as the wedge angle increased. The strut surface temperature decreased as the material thermal conductivity increased. Shi and Wang (2008) studied on an analytic solution of a simplified local thermal non-equilibrium model to optimize a porous structure which consists of two layered media. Their results indicated that the thermal conductivity and porosity of the second layer near the hot side is very important for the hot side temperature. And when the pressure difference is constant, the porosity of the first layer has notable effect on the coolant mass flow rate and when the coolant mass flow rate is constant the porosity of the first layer has a miner effect on the hot surface temperature. Under the condition of constant pressure or constant mass flow rate, the composition and porosity of the second layer have more important effects on the hot surface temperature than the first layer. Most of these previous investigations can be divided into two categories. The first group focused on analyses of the characteristics of the boundary layer, turbulent or laminar flow in transpiration conditions, as heat transfer, skin factor coefficient and flow separation. The second group focused on transpiration cooling effectiveness at high pressure and temperature to simulate practical operational conditions. However, there are few studies of controlling the temperature of hot gas stream and surface to enhance heat transfer by using transpiration cooling. The objective of this study is to investigate the local wall temperature and cooling effectiveness distribution along the surface of a porous flat plate with air as a hot gas stream and water as a coolant to figure out the influence of Reynolds number of hot gas stream and different inlet temperature on heat transfer. Experimental Apparatus A schematic of the experimental facility, which consists of an air blower, flow meter for air and water, a computerized data acquisition system, a test section, a power unit, an absolute and differential manometer and a calibrator thermometer, is shown in Fig. 2. The power unit includes variac and parallel connected air heaters. The test section is composed of a polycarbonate rectangular channel and a porous plate. Details of the test

8 208 Kılıç section are shown in Fig.3. The channel is made of polycarbonate sheet. Dimensions of test section were arranged as 220x880x10 mm (width, length, high). Figure 2. Schematic of experimental apparatus Figure 3. Details of the test section Hydraulic diameter of the channel is 19.1 mm. Then porous plates (high performance polyethylene) were arranged and located over water channel in test section precisely. Porous plate consists of round balls of which particle diameter is 200 μm. Temperatures at the centers of porous

9 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), plate were measured using calibrated T-type thermocouples inserted through 2 mm holes inside the thickness of the porous plate. Calibrated thermocouples were located in the middle of porous plates in this arrangements (6,16,28,38,50,60,72,82 cm). Conventional temperature measurements of the porous plate surface were used to evaluate cooling efficiency of the system. Two electric heaters were used to support enough power to the system to increase air temperature to the needed level. This electric heater was connected parallel and they supplied W power to the system according to power we need. A variac was used to arrange voltage to support needed power for the system. The total power supplied was monitored using digital multimeters for the control of voltage and current. Then flow meter of air and water were added to the system. Air temperatures ranged between 47 C and 77 C. Polystyrene foam board was used to insulate the top side of the channel with a thickness of 250 mm (k = W/(m.K)). The ambient temperatures in the experiments varied between 20 C to 24 C. The system was assumed to be steady state when variations of the wall temperatures and the inlet and outlet fluid temperatures were all within ±0.1 ºC. Data Reduction In this application there is a porous plate in a rectangular channel, this porous plate is wetted by water of which reservoir temperature is T water = 22 ºC. Air enters in the channel with some different temperatures and velocities. For this application dry air was used as a hot gas stream. In this application; all Reynolds numbers were bigger that 3000 so it can be assume that it is a turbulent flow. Bulk temperature of hot air in the channel is calculated as; T avg T (1) airin T 2 surface where T airin is the inlet temperature of air, T surface is the average surface temperature of porous plate. So It can be calculated, as explained by Mills (2001), some physical properties of air by using thermodynamics tables as; ρ is the density of air, cp specific heat of air, γ is the kinematic viscosity, Pr is the Prandtl number and Sc Schmidt number. So Stanton number with zero porosity can be calculated as;

10 210 Kılıç St * (Re) 0.2 (Pr) 2 / 3 (2) And convective heat transfer coefficient with zero-mas-transfer will be; hc *. V. Cp. St air * (3) where ρ is the density of air, V air is air velocity, cp specific heat of air and St* is stanton number. So mass transfer driving force is; B m mev m m 1 s s (4) where m ev is mass fraction of water vapor at channel flow and m s is mass fraction of water vapor at surface. And mass transfer St number can be calculated as; * St mass (Re) 0.2 ( Sc) 2/3 (5) So the zero mass transfer limit conductance will be; g * mass. V air. St * mass (6) where ρ is the density of air, V air is air velocity and St * mass is the mass transfer stanton number. And the mass transfer conductance is; g mass g g * mass * ln(1 mass. B m B m ) where is the the zero mass transfer limit conductance, B m is the mass transfer driving force. So mass flow rate of evaporated water is; (7) m g ev mass. B m (8)

11 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Then heat transfer blowing parameter is; B h m. Cp ev * hc (9) m where ev is the mass flow rate of evaporated water, cp specific heat of air and hc* is convective heat transfer coefficient with zero-mas-transfer. And convective heat transfer coefficient for surface of porous plate is; h c h * c Bh. exp( B h ) 1 (10) So energy balance equation on the surface of porous plate is; m air m.( hairout hairin).( hwaterin hfg hwaterout) water (11) where h airin h waterin m air is the mass flow rate of air, hairout is the outlet enthalpy of air, m is the inlet enthalpy of air, water is the mass flow rate of water, is the inlet enthalpy of water, hwaterout is the outlet enthalpy of water, h fg and is the latent heat of water. Convection heat transfer can be calculated as; conv q m water.( h fg hwaterout hwaterin) hc. A.( Tair Tsurafce) (12) where m h water is the mass flow rate of water, waterin hwaterout h is the inlet enthalpy of fg water, is the outlet enthalpy of water, is the latent heat of water, h c is the convective heat transfer coefficient for surface of porous plate, A is surface area of porous plate, T air is the temperature of air and T surface is the avarage surface temperature of porous plate. Conductıon heat transfer from porous plate to water will be;

12 212 Kılıç cond q m m water.( h h ) waterin surfece (13) h where water surface is the mass flow rate of water, is the enthalpy of water at surface temperature, h waterin is the enthalpy of water at inlet temperature. So cooling efficiency of the porous plate is; T T surface water T T airin airin And cooling efficiency of the system will be; (17) sys T T airin airin T T airout surface (18) Numerical Model FLUENT ANSYS 14.0 simulation program was used for numerical analysis. Channel dimensions were 880 x 220 x 10 mm. At beginning air as a hot gas stream entered the channel at at different temperature. These temperatures were changed according to the heat flux. Air entered the channel at different speeds according to Reynolds Number of hot air. Water as a coolant entered from bottom of the porous plate at 22 ºC. When water reaches the hot side of porous plate it evaporates and it cools the surface by getting latent heat from the surface and mixed with the air stream while cooling air too. By decreasing surface temperature main surface can be protected against high heat flux. Addition to this cool air process can also be used for many industrial applications, space researches and defense applications. Model geometry is shown in Fig.4.

13 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Inlet Channel Outlet Poros Plate Poros Plate Figure 4. Model Geometry We used 220x60x30 ( elements) mashes for our applications. We prepared mash structure according to flow conditions. In order to get more precise numerical results we increased mash numbers in some region as surface of porous plate, inlet and outlet of channel and near wall regions of channel. Minimum orthogonal quality is 1 and minimum aspect ratio is It is observed that numerical geometry was independent from sweep number and cell number when sweep number was 600 and cell number was 220x60x30. The calculation continued for about 600 more iteration after the maximum residuals for all the dependent variables were less than 10-5, the overall energy and mass balances were less than Boundary conditions used in numerical study are shown in Table 1. Table 1. Boundary conditions Air Inlet U V W T U U airinlet 0 V W 0 T Tairinlet Water U 0 inlet V 0 W Wwaterinlet Porous U 0 V 0 0 Plate U V W Outlet T Tout x x x Front and T U = 0 V = 0 W =0 0 back wall y T T waterinlet q q conv W. T z Top wall U = 0 V = 0 W =0 0

14 214 Kılıç The continuity, Reynolds averaged momentum and time averaged energy equations governing 3-dimensional steady, flow of air with constant properties used for turbulent solutions can be written in the Cartesian coordinate system as follows: Continuity equation: (19) Momentum equation: (20) Energy equation: (21) The Navier-Stokes equations were solved with several turbulence models including the Standard k-ε model, RNG k-ε model, Realizable k-ε model and SST k-ω model to predict the thermal and fluid characteristics of porous plate subjected to transpiration cooling. The values of y+ were less than 2.5 along the walls. Validation of turbulence model with experimental results was shown in Fig.5. It was seen that RNG k-ε model best represents the general trends of the hot air temperatures and surface temperature of porous plate. 0 i i x U j i i j j i i j i j i u u x U x U x x P x U U u T c x T k x x T U c i p i i i i p

15 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Standart k- Realizable k- RNG k k- Exp. T ( o C) ,0 0,2 0,4 0,6 0,8 X (m) Figure 5. Validation of turbulence models with experimental results After forming a model geometry and mash structure, independent from sweep numbers and cell numbers, boundary conditions were applied for heat transfer from the surface with transpiration cooling. Results and Discussion In this section, experimental and numerical results were presented for different Reynolds number and temperature of hot gas stream. Numerical results were also compared and verified with experimental results. Effect of Reynolds number of hot gas stream (air) Experiments were conducted for different Reynolds number for T inlet = 77ºC, particle diameter D p = 200 μm and ṁ water = 0, kg/s. Velocity profile of hot gas stream (air) were chosen in turbulent flow regime with low Reynolds number because it is planned to see effect of porous plate and coolant on heat transfer on the porous plate precisely. For different Reynolds number, measured temperature values of surface were shown in Fig.6.

16 216 Kılıç Re= 3035 Re= 3200 Re= 3300 Re= T ( o C) ,0 0,2 0,4 0,6 0,8 X (m) Figure 6. Surface temperature for different Reynolds number It can be seen that effect of Reynolds number can be seen clearly at the inlet region of porous plate. But by cooling of air along the porous causes a gradually decrease on difference of surface temperature up to the end of porous plate. Temperature difference is minimum at the outlet region of porous plate. So there is not a significant difference on surface temperature and cooling efficiency for different Reynolds number we used at the end of porous plate. Because Reynolds numbers were chosen in a narrow gap to see effect of Reynolds number near to optimum value. Temperature values changed between 77 ºC and 27 ºC. Surface temperature is decreasing sharply up to x=0,28 m but after that point there is not a prominent difference on surface temperature. Cooling efficiencies of porous plate for different Reynolds number were shown in Figure 7.

17 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Efficiency (%) Re=3035 Eff. Re=3200 Eff. Re=3300 Eff. Re=3580 Eff. 20 0,0 0,2 0,4 0,6 0,8 X (m) Figure 7. Cooling efficiency of porous plate for different Reynolds number Cooling efficiency of porous plate is changing between 38 and 90 % along the porous plate. As surface temperature, cooling efficiency increases up to x=0,28 m after this point it changes slightly. So we could observe that increasing Reynolds number causes an increase on surface temperature and a decrease on cooling efficiency of porous plate. Cooling efficiencies of system for different Reynolds number were shown in Figure System Eff. 60 Efficiency (%) Re Numbers Figure 8. Cooling efficiency of system for different Reynolds number Increasing Reynolds number causes a slight decrease on efficiency of the system. Increase of 15.2% on Reynolds number causes a decrease of efficiency of the system of 6.9%. Decreasing trend of the efficiency of the

18 218 Kılıç system also increases gradually between Re= 3035 and 3580.The reason of this decrease on the efficiency of the system is that increasing air velocity (while increasing Reynolds number) causes a decrease on heat convection between hot air and water on the surface of porous plate. Numerical analysis is conducted for different Reynolds number (Re=3035, 3200, 3300, 3580, 6500 and 9430) at T airinlet = 77 ºC, particle diameter D p =200 μm and ṁ water =0, kg/s. Temperature contours for Re=3035 and 9430 were shown in Fig o C 74 o C 72 o C 69 o C 66 o C 63 o C 61 o C 58 o C 53 o C 52 o C 50 o C 47 o C 44 o C 41 o C 39 o C 33 o C 30 o C 27 o C 25 o C 23 o C 22 o C (a) (b) Figure 9. Temperature contours for (a) Re=3035 and (b) Re=9430 It can be seen that surface temperature is higher at regions near to the walls, because fluid velocity near to the walls is low. Increasing Reynolds number 3 times causes a slight increase on surface temperature, 4 ºC, of 1.35 % on average surface temperature. There is a slight reduce on surface temperature from inlet to outlet of the channel. Variation of surface temperatures for different Reynolds number is shown in Fig.10. Effect of Reynolds number of hot air can be detected easily at the inner region of porous plate for higher Reynolds number according to lower Reynolds number.

19 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Re= 3035 Re= 3200 Re= 3300 Re= 3580 Re= 6500 Re= 9430 T ( o C) ,0 0,2 0,4 0,6 0,8 X (m) Figure 10. Variation of surface temperatures for different Reynolds number Effect of Temperature of hot gas stream (air) Experiments were conducted for different temperature of hot gas stream for Re=3300, particle diameter Dp=200 μm and ṁ water = 0, kg/s. Inlet temperature of air was changed by changing voltage of current on variac. Measured surface temperatures for different air inlet temperature were show in Fig T airinlet = 47 o C T airinlet = 57 o C T airinlet = 67 o C T airinlet = 77 o C 50 T ( o C) ,0 0,2 0,4 0,6 0,8 X(m) Figure 11. Surface temperatures for different air inlet temperature

20 220 Kılıç It was observed that decreasing of inlet temperature caused a prominent decrease on surface temperature especially up the middle of the porous plate. Up to the end of the porous plate temperature difference decreases and in that region there is not a significant effect on surface temperature for different inlet temperature. Cooling efficiency of porous plate for different air inlet temperature was shown in Fig Efficiency (%) T input = 47 o C 20 T input = 57 o C T input = 67 o C T input = 77 o C 0 0,0 0,2 0,4 0,6 0,8 X (m) Figure 12. Cooling Efficiency of porous plate for different air inlet temperature It is obtained that decreasing inlet temperature caused a prominent increase on cooling efficiency of porous plate generally. Effect of hot gas stream temperature can be seen along the plate. Between inlet and outlet; surface temperature can be decreased of 13,4% (from 77 to 30 ºC). Beyond that point temperature decrease was so small. Difference of cooling efficiency of porous plate between 77 ºC and 47 ºC is approximately 10 %. Increasing inlet temperature causes an increase on cooling efficiency. Cooling efficiency of the system was shown in Fig. 13. Increasing air inlet temperature does not cause a significant increase on cooling efficiency of the system. Increasing air inlet temperature of 9.4 % causes

21 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), only an increase on efficiency of the system of 0.7%. The reason of this is increasing air inlet temperature increases not only heat convection from the porous plate but also temperature of surface of porous plate and outlet temperature of the air. 65 System Eff. 60 Efficiency (%) T airinlet Figure 13. Cooling Efficiency of the system for different air inlet temperature Numerical analysis were conducted for different temperature of hot gas stream for Re=3300, particle diameter D p =200μm and ṁ water =0, kg/s. Temperature contours for T inlet =47 ºC and 77 ºC were shown in Fig.14. There is not a significant effect of inlet air temperature on surface temperature at the end of porous plate. Difference between inlet and outlet temperatures of surface increase by increasing inlet air temperature. Variation of surface temperature for different inlet temperatures is shown in Fig.15.

22 222 Kılıç 47 o C 46 o C 45 o C 43 o C 42 o C 41 o C 40 o C 38 o C 37 o C 36 o C 35 o C 33 o C 32 o C 31 o C 30 o C 28 o C 27 o C 26 o C 25 o C 23 o C 22 o C (a) 77 o C 74 o C 72 o C 69 o C 66 o C 63 o C 61 o C 58 o C 53 o C 52 o C 50 o C 47 o C 44 o C 41 o C 39 o C 33 o C 30 o C 27 o C 25 o C 23 o C 22 o C (b) Figure 14.Temperature contours for (a) T inlet = 47 ºC and (b) T inlet =77 ºC 80 T airinlet = 47 o C 70 T airinlet = 57 o C T airinlet = 67 o C 60 T airinlet = 77 o C T ( o C) ,0 0,2 0,4 0,6 0,8 X (m) Figure 15. Variation of surface temperature for different inlet temperatures

23 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Validation of numerical results with experimental results After obtaining experimental and numerical results validation of numerical results was shown with graphs which show approximation of numerical results with experimental results. Numerical results generally have a good approximation with experimental results. It was also observed that 4th and 5th group thermocouple showed higher temperature according to numerical result. It is thought that the reason of this is dry point of porous plate. The most temperature difference between experimental and numerical results is 5.2 ºC or 1.5 %. Comparison of experimental and numerical values of surface temperature for different Reynolds number and different air inlet temperatures are shown in Fig.16, 17. T ( o C) Re= 3035 Exp. Re= 3200 Exp. Re= 3300 Exp. Re= 3580 Exp. Re= 3035 Num. Re= 3200 Num. Re= 3300 Num. Re= 3580 Num ,0 0,2 0,4 0,6 0,8 X (m) Figure 16. Comparison of experimental and numerical values of surface temperature for different Reynolds number 80 T airinput = 47 o C Exp. 70 T airinput = 57 o C Exp. T airinput = 67 o C Exp. 60 T airinput = 77 o C Exp. T airinput = 47 o C Num. T ( o C) 50 T airinput = 57 o C Num. T airinput =67 o C Num. 40 T airinput = 77 o C Num ,0 0,2 0,4 0,6 0,8 X (m) Figure 17. Comparison of experimental and numerical values of surface temperature for different inlet temperatures

24 224 Kılıç Conclusions The present study is focused on investigation of heat transfer from a porous plate by cooling of air and surface with transpiration cooling. Surface temperature of porous plate were measured and cooling efficiency of system were calculated for different Reynolds number of hot gas stream (air) (Re= 3035, 3200, 3300, 3580, 6500, 9430), temperature of air (T air = 47, 57, 67, 77 ºC). The following conclusions can be drawn from the experimental and numerical results; Effect of Reynolds number can be seen clearly at the inlet region of porous plate. Cooling efficiency of porous plate is changing between 38 and 90 % along the porous plate. So increasing Reynolds number causes an increase on surface temperature and a decrease on cooling efficiency of porous plate. Increasing Reynolds number causes a slight decrease on efficiency of the system. Increase of 15.2% on Reynolds number causes a decrease of efficiency of the system of 6.9%. Decreasing of inlet temperature causes not only a significant decrease on surface temperature but also a slight increase on cooling efficiency of the system. Between inlet and outlet; surface temperature can be decreased of 13.4% (from 77 to 30 ºC). Beyond that point temperature decrease was negligible. Difference of cooling efficiency of porous plate between 77 ºC and 47 ºC is approximately 10 %. Increasing air inlet temperature does not cause a significant change on cooling efficiency of the system. Increasing air inlet temperature of 9.4 % causes only an increase on efficiency of the system of 0.7%. Numerical results generally have a good approximation with experimental results. The most deference between experimental and numerical results is 5.2 ºC or 1.5%. Research areas for future investigations may be effect of particle diameter of porous plate on heat convection, effect of different coolant fluid (nonufluids etc.) on heat transfer from porous plate, effect of shape of porous plate on heat transfer and understanding characteristics of thin film occurred on porous plate for heat transfer and fluid flow. Acknowledgement The financial support of this study by the Scientific Research Council of Turkey (TUBITAK), with the program of postdoctoral

25 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), scholarship (2219) and University of California Los Angeles Post Doctorate Program (UCLA/USA), is gratefully acknowledged. References Andoh, Y.H., Lips, B. (2003). Prediction of porous walls thermal protection by effusion or transpiration cooling. An analytic approach. Applied Thermal Engineering, 23, Arai, M., Suidzu, T. (2012). Porous ceramic coating for transpiration cooling of gas turbine blade. J. Thermal Spray Technology, 22, Cerri, G., Giovannelli, A., Battisti, L., Fedrizzi, R.(2007). Advances in effusive cooling techniques of gas turbines. Applied Thermal Engineering, 27, He, F., Wang, J., Xu, L., Wang, X.(2013). Modeling and simulation of transpiration cooling with phase change. Applied Thermal Engineering, 58, He, F., Wang, J.(2014). Numerical investigation on critical heat flux and coolant volume required for transpiration cooling with phase change. Energy Conversion and Management, 80, He, S., Guan, Z., Gurgenci, H., Hooman, K., Lu Y., Alkhedhair, A.M.(2014). Experimental study of film media used for evaporative pre-cooling air. Energy Conversion and Management, 87, Huang, Z., Zhu, Y., Xiong, Y., Jiang, P.(2014). Investigation of transpiration cooling for sintered metal porous struts in supersonic flow. Applied Thermal Engineering, 70, Jiang, P.X.,Yu, L.,Sun, J.G.,Wang, J.(2004). Experimental and numerical investigation of convection heat transfer in transpiration cooling. Applied Thermal Engineering, 24, Langener, T., Wolserdorf, J., Selzer, M., Hald, H.(2012). Experimental investigations cooling applied to C/C material. Int. J. Thermal Science, 54, Liu, Y.Q., Jiang, P.X., Jin,S.S.,Sun, J.G. (2010).Transpiration cooling of a nose cone by various foreign gases. Int. Journal of Heat and Mass Transfer, 53, Liu, Y.Q., Jiang, P.X., Xiong, Y.B.Wang, Y.P. (2013). Experimental and numerical investigation of transpiration for sintered porous flat plates. Applied Thermal Engineering, 50,

26 226 Kılıç Liu, Y.Q., Xiong, Y.B., Jiang, P.X., Wang, Y.P., Sun, J.G.(2013). Effects of local geometry and boundary conditions on transpiration cooling. Int. J. Heat and Mass Transfer, 62, Mills, A.F., (2001). Mass Transfer. University of Los Angeles California, Los Angeles. Polezhaev, J.,(1997). The transpiration cooling for blades of high temperatures gas turbine. Energy. Convers. Manage., 38, Shi, J., Wang, J.(2008). Optimized structure of two layer porous media with genetic algorithm for transpiration cooling. Int. J. Thermal Science, 47, Trevino, C., Medina, A. (1997). Analysis of transpiration cooling of a thin porous plate in a hot laminar convective flow. Eur. J. Mech., 2, Tsai, G., Lin, Y.C., Wang, H.W., Lin, Y.F., Su, Y.C., Yang, T.J. (2009). Cooling transient in a sudden-expansion channel with varied rates of wall transpiration. Int. J. Heat and Mass Transfer, 52, Tsai, Y.Y, Lee, C.H.(2014a). Experimental study of evaporative heat transfer in sintered powder structure at low superheat levels. Experimental Thermal and Fluid Science, 52, Tsai, Y.Y, Lee, C.H.(2014b). Effect of sintered structural parameters on reducing the superheat level in heat pipe evaporators. Int. J. Thermal Science, 76, Wang, j., Zhao, L., Wang, X., Ma, J., Lin, J. (2014). An experimental investigation on transpiration cooling of wedge shaped nose cone with liquid coolant. Int. J. Heat and Mass Transfer, 75, Zhao, L., Wang, J., Ma, J., Lin, J., Peng, J., Qu, D., Chen, L.(2014). An experimental investigation on transpiration cooling under supersonic condition using a nose cone. Int. J. Thermal Science, 84,

27 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), Extended Summary Askerî Sistemler için Yeni Bir Soğutma Tekniği; Terleme ile Soğutma Giriş Terleme ile soğutmanın dikkat çekici soğutma tekniklerinden biri olduğu bilinmektedir. Bu yöntem; hipersonik hava araçlarında, sıvı yakıtlı roketlerde, gaz türbin kanatçıklarında, su oksidasyon teknolojisinde, atmosfere giriş yapan uzay mekiklerinin burun kısmında ve gövdelerinde, çeşitli mühimmatlarda ve roketlerde ve birçok askerî uygulamada kullanılabilmektedir. Terleme ile soğutma süreci; iki farklı ısı transferi mekanizmasını içerir; katı geçirgen yüzeyden iletim ve taşınım. Sonuç olarak soğutucu akışkan gözenekli plaka yüzeyinde ince bir film oluşturarak, hem yüzeyi korur hem de taşınımla ısı transferi ile sıcak/soğuk akışkana ısı alır/verir. Bu konuda literatürdeki çalışmalar iki grupta toplanabilir; Birinci grup; terleme ile soğutma tekniğinde sınır tabaka, türbülanslı ve laminer akış uygulamalarına odaklanmış, ikinci grup ise; pratik uygulamalar için soğutma verimi üzerine odaklanmıştır. Ancak terleme ile soğutma tekniği kullanılarak; yüzeyin yanı sıra sıcak akışkanın da (hava) soğutulduğu bir süreçte farklı parametrelerin etkisinin incelendiği, ısı transferini ve dolayısı ile sistem verimini artırmaya yönelik çok az çalışma mevcuttur. Bu çalışmanın amacı; havanın sıcak akışkan ve suyun soğutucu akışkan olarak kullanıldığı (birçok pratik uygulamada kullanılabilecek olduğu gibi) bir sistemde yerel gözenekli plaka yüzey sıcaklıklarının ve sistem soğutma veriminin incelenerek, sıcak akışkan Reynolds sayısının ve farklı giriş sıcaklıklarının ısı transferine etkilerini ortaya koymaktır. Deney Setleri Deney seti; fan, hava ve su için debi ölçer, bilgisayarlı veri kayıt sistemi, bir test kısmı, güç ünitesi, mutlak ve diferansiyel manometre, bir kalibratör termometreden oluşmaktadır Güç ünitesi; bir varyak ve paralel bağlanmış iki hava ısıtıcısından oluşmaktadır. Test kısmı; polikarbon bir kanal ile yüksek performanslı polietilen malzemeden yapılmış 200 micrometre parçacık çaplı (küre şeklinde küçük toplardan oluşan) gözenekli bir plakadan oluşmaktadır. Sıcak akışkan ısısı iki adet elektrikli ısıtıcı kullanılarak sağlanmıştır. Varyak; sisteme ihtiyaç duyduğu ısı değerlerini sağlayacak voltaj değerlerinin ayarlanmasında kullanılmıştır. İhtiyaç

28 228 Kılıç duyulan toplam güç değerini sağlayacak voltaj ve akım değeri dijital görüntüleme sistemi ile okunmuştur. Daha sonra ise hava ve su debi ölçerleri sisteme dâhil edilmiştir. Hava giriş sıcaklığı; 47 C-77 C aralığındadır. Deneylerin yapıldığı ortam sıcaklıkları 20 C-24 C aralığındadır. Sistem sürekli haldedir ve sıcaklık hata ölçüm aralığı ±0,1 ºC dir. Sayısal Model Sayısal analiz için FLUENT ANSYS 14,0 simülasyon programı kullanılmıştır. Terleme ile soğutma sürecinde; geçirgen plaka yüzeyindeki ısıl ve akış karakteristikleri; Navier-Stokes denklemlerinin Standard k-ε, RNG k-ε, Realizable k-ε, ve SST k-ω modellerinde çözülmesiyle incelenmeye çalışılmıştır. Duvara yakın bölgelerde y+ değerleri 2,5 den azdır. RNG k-ε modelinin gözenekli plaka yüzeyindeki sıcaklık ve akış özelliklerini diğer modellere göre daha iyi temsil edebildiği görülmüştür. Sonuçlar Bu çalışmanın amacı; sıcak akışkan olarak havanın ve soğutucu olarak suyun kullanıldığı durumda, hem sıcak akışkanın hem de yüzeyin soğutulması maksadıyla, gözenekli bir plakadan olan ısı transferinin terleme ile soğutma tekniği kullanılarak incelenmesidir. Isılçiftler yardımıyla ölçülen hava sıcaklığı ve gözenekli plaka yüzey sıcaklıkları; gözenekli plaka soğutma verimini ve sistem soğutma verimini hesaplamak için kullanılmıştır. Çalışma parametreleri; sıcak akışkan kanal hızına ve sıcaklığına bağlı olarak hesaplanan Sıcak akışkan Reynolds sayısı (Re= 3035, 3200, 3300, 3580, 6500, 9430) ve sıcak akışkan sıcaklığıdır (T hava = 47, 57, 67, 77 ºC). Deneysel ve sayısal çalışmalardan elde edilen sonuçlar aşağıda sunulmuştur; Reynolds sayısının artırılmasının plaka yüzey sıcaklığının yükselmesine, gözenekli plaka soğutma veriminin azalmasına sebep olduğu tespit edilmiştir. Reynolds sayısında % 15,2 lik bir artış, sistem soğutma veriminde % 6,9 luk bir azalmaya sebep olmuştur. Plaka boyunca gözenekli plaka soğutma verimi % aralığında değişmiştir. Hava giriş sıcaklığının azaltılması hem plaka yüzey sıcaklığında önemli bir azalmaya sebep olmuş hem de sistem soğutma veriminde kısmi bir artışa sebep olmuştur. Kanal giriş ve çıkışı arasında yüzey sıcaklığı % 13,4 oranında (77-30 ºC) azaltılabilmiştir. Sıcak hava giriş sıcaklığı 47 ºC-77 ºC aralığında arttırıldığında gözenekli plaka soğutma verimi

29 Savunma Bilimleri Dergisi, Mayıs 2016, 15 (1), yaklaşık % 10 oranında artmıştır. Giriş hava sıcaklığında % 9,4 lük bir artış sistem soğutma veriminde yalnızca % 0,7 lik bir artışa sebep olmuştur. Sayısal sonuçlar deneysel sonuçlarla genel olarak iyi bir uyum içerisindedir. Deneysel ve sayısal sonuçlar arasındaki en büyük fark; 5,2 ºC ya da %1,5 dir. Gözenekli plaka parçacık çapının, farklı geometrideki gözenekli plakaların, farklı tip soğutucuların (nanoakışkanlar vb.), ısı transferine etkisinin ve plaka yüzeyinde oluşan ince film tabakasının ısıl ve akış karakteristiklerinin incelenmesinin gelecekte yapılacak araştırmalarda faydalı olabileceği değerlendirilmiştir.