- Email: [email protected]
The experimental study of nanofluids boiling crisis on cylindrical heaters
International Journal of Thermal Sciences 116 (2017) 214e223
Contents lists available at ScienceDirect
International Journal of Thermal Sciences journal homepage: www.elsevier.com/locate/ijts
The experimental study of nanofluids boiling crisis on cylindrical heaters A.V. Minakov a, b, c, *, M.I. Pryazhnikov a, b, c, D.V. Guzei a, c, G.M. Zeer a, V. Ya. Rudyak c a
Siberian Federal University, Krasnoyarsk, Russia Kutateladze Institute of Thermophysics, SB RAS, Novosibirsk, Russia c Novosibirsk State University of Architecture and Civil Engineering, Russia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 April 2016 Received in revised form 26 January 2017 Accepted 27 February 2017
The paper deals with the experimental study of nanofluids boiling on cylindrical heater. The studied nanofluids were prepared on the basis of distilled water and nanoparticles of silicon, aluminum and iron oxides as well as diamond. The volume concentration of nanoparticles varied from 0.05 to 1%. The nanoparticles diameter ranged from 10 to 100 nm, the diameter of the heater was varied from 0.1 to 0.3 mm. It is revealed that the use of nanofluids provides a several-fold increase of the critical heat flux. However, critical heat flux in nanofluids depends on the material and size of nanoparticles as well as on the diameter of used cylindrical heater. It is shown that the critical heat flux increases with increasing the nanoparticles size, while it decreases with increasing the heater diameter. It was revealed for the first time that the critical heat flux depends on the boiling process duration. © 2017 Elsevier Masson SAS. All rights reserved.
Keywords: Nanofluid Nanoparticles Boiling crisis Critical heat flux (CHF) Deposition of nanoparticles Heater surface microstructure
1. Introduction Boiling, as one of the most effective and efficient ways of heat transfer, is used in various engineering applications. Therefore, the enhancement of the critical heat flux in boiling has been the subject of numerous studies carried out in the last century [1e4]. An intensive research on the possible use of nanofluids in boiling-related applications has started just recently. Apparently, works [5,6] should be considered as the first investigations of nanofluids boiling heat transfer and crisis. The first work dealt with the boiling of water-based silicon dioxide and aluminum oxide nanofluids on a square heater with a characteristic size of 10 mm. The authors observed a sharp increase in the critical heat flux (CHF) caused by the presence of nanoparticles. At the pressure of about 0.198 bar, the increase in CHF reached more than 200% when using water-based Al2O3 nanofluid with nanoparticle concentration of 5$104 wt%. It was also noted that the supplement of nanoparticles to water increases the size of bubbles, though reduces the frequency of their detachment. It is unclear, however, how these changes are related to the observed increase in CHF.
* Corresponding author. Siberian Federal University, Krasnoyarsk, Russia. E-mail address: [email protected] (A.V. Minakov). http://dx.doi.org/10.1016/j.ijthermalsci.2017.02.019 1290-0729/© 2017 Elsevier Masson SAS. All rights reserved.
In Ref. [6], a nucleate pool boiling was studied on the surface of cylindrical heating element 20 mm in diameter in the water-based alumina nanofluid. Contrary to work [5], it was shown that the presence of nanoparticles degrades the boiling characteristics (heat transfer coefficient), and this deterioration increases with increasing nanoparticle concentration. A similar phenomenon was observed in a later study [7], where smaller heaters with an outer diameter ranged from 4.5 to 6 mm were used. The authors explained the deterioration in heat transfer by the changes of heaters surface characteristics. They claim that during boiling of nanofluids, the surface became smoother due to the deposition of nanoparticles at the nucleation sites. The higher the concentration, the smoother the surface, which in turn leads to a greater reduction of heat transfer coefficient. This explanation is not consistent with observations [8], concerned with the study of water-based alumina nanofluids boiling on a square surface with the characteristic size of 100 mm at high heat fluxes. It was revealed that after boiling, the surface roughness increased with increasing the nanoparticle concentration. Vassallo et al. [9] studied boiling of water-based silicon dioxide nanofluid on the horizontal nickel-chromium wire with a diameter of 0.4 mm at atmospheric pressure. Similar to work [5], a two-fold increase in CHF was observed.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Tu et al. [10] conducted boiling experiments in water-based alumina nanofluid in a relatively large volume with rectangular heating surface of 26 40 mm2 at atmospheric pressure. A significant increase in both heat transfer and CHF was obtained in the nucleate boiling mode. For example, compared to pure water boiling, an increase in heat transfer coefficient by more than 60% was recorded while investigating highly diluted aqueous nanofluids. A significant reduction in the maximum wall temperature of the heater as well as a four-fold increase in the number of active bubble nucleation sites was found in comparison with pure water boiling. The boiling heat transfer in water-based nanofluid with gold nanoparticles was studied in Ref. [11]. A significant increase in heat transfer coefficient was found during boiling at atmospheric pressure on the heater with the diameter of 100 mm. In particular, the authors managed to increase the heat transfer coefficient of nanofluid containing 0.001 wt % of gold nanoparticles by 21% at the heat flux of 45 kW/m2. The increase in boiling heat transfer coefficient was also accompanied by the growth of critical heat flux with increasing nanoparticle concentration. Most researchers observe the deposition of nanoparticles on the heater surface in the course of boiling. Apparently [12], was the first work, where the increase in CHF was explained by the presence of nanoparticles depositions (see also [13]). It was shown that the wettability of heater surface covered by deposited nanoparticles was much higher than that of a pure surface. The key role of nanoparticles deposition on the heater surface in the boiling crisis and the effect of surface wettability on the CHF were also observed in the later works [14e28]. The effect of nanoparticle concentration on CHF has been studied well enough to date. It is revealed that in most cases, increasing concentration of nanoparticles increases CHF, though in some cases there was a decrease in CHF at high concentrations of nanoparticles [15]. In addition, the effect of heater surface area on the CHF was observed in Refs. [17,18]. It is shown in Ref. [18] that increasing the size of the heater slows down the enhancement of CHF in nanofluids. The effect of pressure on CHF was studied in Ref. [19], where it was revealed that the gain in CHF the value due to use of nanofluids instead of pure fluids decreases with the increase of pressure. The data on the effect of particle size on CHF are rather contradictory. Thus, Moreno et al. [20] claimed that CHF does not depend on the size of alumina nanoparticles within the particle size range from 69 to 346 nm. Similar conclusions were made in Ref. [21]. At the same time, it was found in Ref. [15] that CHF decreases with increasing particle size in nanofluids containing silver particles. On the contrary, in Refs. [22,23] it was shown that CHF increases with increasing particles size. In many studies, it was observed that the effect of increased CHF was retained even in boiling of pure fluids on the heaters covered by nanoparticles [18,24,25]. Among the recent works, we should note theoretical papers and surveys [26e29], which are dealing with detail consideration of various mechanisms leading to enhancement of CHF in boiling nanofluids. However, the decisive conclusions on these mechanisms have not yet been made. Thus, despite a large number of experimental data, they are quite contradictory and do not allow comprehending clearly the mechanisms of CHF variations in the boiling process of nanofluids. In order to remove at least some of these contradictions, this paper provides with the systematic experimental study of saturated pool boiling of nanofluids on a cylindrical heater. Our main objective is to understand the dependence of CHF on the nanoparticle size and material, heat diameter, and nanoparticles deposition on its surface.
215
2. Description of the experimental setup The scheme of experimental setup is presented in Fig. 1. The studied fluid in amount of 300 ml was placed in a high sealed glass flask (4) with the diameter of 8 cm. A nickel-chromium wire heater with the length L ¼ 34 mm and the diameter dh ¼ 0.2 mm was emerged into the flask filled with fluid. The wire was fixed by copper bus leads with the cross section of 10 2 mm to supply voltage. The heater was energized by means of GPS-6030D power source. The applied voltage was recorded using a GDМ-78261 voltmeter (“V” in Fig. 1). Load amperage in the circuit was measured by the APPAe109N ammeter (“A” in Fig. 1). The flask with the test fluid was sealed, so that the condensate formed in the upper part of the flask dripped back into the flask, maintaining saturation conditions in the working chamber. The fluid temperature in the flask was controlled by means of a chromel-copel thermocouple, which was connected to the TRM-200 temperature meter (5). The tail end of the thermocouple was located at the same level with the nickelchromium heater at a distance of 2 cm from it. The flask with the test fluid was placed in a water bath (3) with the constant temperature, which was about 0.5 C below the boiling point and maintained by means of electric heater (2). Thus, in this paper we investigate the boiling close to saturation conditions. The temperature control in a water bath was conducted using another chromel-copel thermocouple connected to the TRM200 m. After placing the flask into a water bath, the temperatures in the flask and water bath equalized with time. Then the nickelchromium heater was energized, and the heat flux density was measured. The programmable current power supply allowed increasing the heater voltage with a fixed predetermined step. Thus, it was possible to control the heat flux growth rate and, thereby, to fix the onset of boiling crisis. The boiling heat flux density on the heater was determined by the ratio: q ¼ Q =S ¼ IU=S, where S ¼ PdL is the lateral surface area of the heater between the current-carrying wires, Q is the heat flux released by the heater, I is the electric current in the heater circuit,
Fig. 1. The scheme of experimental setup.
216
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
U is the voltage drop in the heater. Heat generated at the lead wires was negligible. The total error in determining the heat flux density was about 2%. Studying the boiling processes requires also the knowledge of heater surface temperature. For the nickel-chromium, the dependence of resistance R on temperature within the investigated temperature range is close to linear R ¼ R0 ½1 þ AðT T0 Þ. Thus, by measuring the resistance of the heater and knowing the temperature coefficient of resistance A, we can find the average heater temperature: T ¼ T0 þ ðR=R0 1Þ=A. Here R0 is the heater resistance at the temperature T0 . To determine the temperature coefficient of nickel-chromium electrical resistance, we have performed special series of resistance measurements of the used nickelchromium wire. It was found that A ¼ 0.00012 1/K. The total error in determining the heater temperature was about 3%. The nanofluids were prepared based on distilled water and nanoparticles of aluminum, silicon, iron (III) oxides as well as nanodiamonds. The volume concentration of particles 4 ranged from 0.05 to 1%. For preparation of nanofluids we used the standard two-step process. After adding the necessary quantity of nanopowder to water, the obtained nanofluid was first thoroughly mixed mechanically. After that, it was placed in an ultrasonic disperser for half an hour to destruct the particles conglomerates. No surfactants were added to the nanofluid. The diameter of the used nanoparticles dp ranged from 10 to 100 nm. The following particle sizes were used: SiO2 (25, 40, 100 nm), Al2O3 (10, 50, 100 nm), Fe2O3 (95 nm), and the diamond particles (5 nm). The first three types of nanopowder were received from “Plazmoterm” company, while nanodiamonds were acquired from FGUP Altay. Measurement of nanoparticles distribution in the fluid in terms of their size was carried out by means of CPS Disk Centrifuge DC24000. Besides measuring the critical heat flux we have investigated also the heaters surface structure after boiling. It was done by using transmission electron microscope JEM-2100 (JEOL, Japan) equipped with an energy dispersive spectrometer Oxford Inca x-sight. 3. Measurement results
Fig. 2. Boiling curves in distilled water and nanofluids with different concentrations of SiO2 particles.
dependence of the relative critical heat flux q=qw (where qw is CHF for water) on the nanoparticle volume concentration for three water-based nanofluids with particles of silicon dioxide (25 nm), iron oxide (95 nm), and diamond (5 nm). The values of critical heat flux presented here were obtained by averaging the data of five independent measurements. The CHF for water was equal to qw ¼ 1:32 MW/m2, which is consistent with the data [12,13]. The analysis of the boiling heat flux behavior in nanofluids shows that at low particle concentrations it increases drastically. However, the growth rate decreases with increasing particle concentration. We can expect that starting from a certain particle concentration CHF will no longer continue to increase. The dependence of the relative critical thermal load q=qw on the particle concentration shown in Fig. 3 significantly differs for the three nanofluids considered. However, in all cases this dependence can be described by the equation
3.1. The dependence of CHF on the nanoparticle concentration The first series of experiments was carried out to determine the dependence of CHF on the nanoparticle concentration. Typical results are presented in Fig. 2 for the water-based nanofluid with SiO2 particles at an average particle size dp of 25 nm and the heater diameter dh of 0.2 mm. The figure shows the dependence of heat flux density on the superheat DT ¼ Tw Tf (where Tw is the heater surface temperature and Tf is the fluid temperature). Even at the nanoparticles concentration of 0.05%, we observed almost two-fold increase in CHF for nanofluid in comparison with that for pure water. This difference in CHF increases with increasing concentration of nanoparticles. At the nanoparticle concentration of 0.5% it becomes almost three-fold. After reaching the critical heat load there was a transition from nucleate boiling to film boiling mode. At that, the heater either burned out instantly or, in very rare cases, continued generating heat in film boiling mode. This a behavior is shown in Fig. 2 for water (see the growth of CHF, which starts at the temperatures above 700 C). In this case, we failed to reach film boiling mode in nanofluids, since the heater was almost always burned out because the thermal load corresponding to CHF in nanofluids during the transition to film boiling mode was 2e3 times higher than that in pure water. The increase in CHF with increasing volume concentration of particles was observed for all nanofluids. Fig. 3 shows the
q=qw ¼ 1 þ a4b
(1)
The constants in equation (1) in general depend on several
Fig. 3. The dependence of the relative critical heat flux on the nanoparticle volume concentration.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
parameters, although for the studied nanofluids they can be taken as follows: a ¼ 2.169, b ¼ 0.1903 for SiO2 particles; a ¼ 1.125, b ¼ 0.4117 for diamond particles; and a ¼ 0.8595, b ¼ 0.3292 for particles of iron oxide Fe3O4. In the majority of works concerned with the nanofluids boiling crisis it is noted that the main reason for increasing of CHF is the nanoparticles deposition on the heater surface. The microstructure of the 0.2 mm heater surface sample and nanofluid coating containing SiO2 particles with an average size of 25 nm is shown in Fig. 4 for different particle concentrations. In all the cases, the boiling process duration was 25 min. It can be seen that the heater surface is covered by fairly dense deposits. With increasing nanoparticle concentration in the fluid at other conditions being equal, the deposits thickness and the proportion of deposited area increase. Thus, the critical heat flux is directly related to deposits of nanoparticles on the heater surface. The fact that these deposits are composed of individual nanoparticles is clearly seen from Fig. 5f at a maximum magnification of 30,000 times. The electron microscopy has shown that at high nanoparticle concentrations, the thickness of deposits reaches about 5e10 mm and further slightly varies with time and with increasing nanoparticle concentration. The analysis of the surface microstructure allows us to conclude that starting from a certain time moment, there exists a dynamic equilibrium between the number of particles deposited on the heater and those washed back into the bulk of fluid. Apparently, it explains the noted slowdown in the growth of CHF and its ultimate saturation caused by increase in particles concentration. It should be noted that the deposits on the heater surface are not solid. The bulk of formed deposits consist of developed microchannels network with a thickness of 5e10 mm and a length of 10e50 mm.
217
3.2. The dependence of CHF on the nanoparticle size The comparison presented in Fig. 3 shows a significant dependence of the relative CHF on nanoparticle volume concentration. However, the three curves shown there are obtained for nanofluids with different particles (in particular, they have different sizes). So far, there are no indisputable data on the dependence of critical heat flux on the nanoparticle size. To ascertain the nature of this dependency we have carried out special series of measurements. Two water-based nanofluids with silica and alumina particles of different size ranged from 10 to 100 nm were considered. The diameter of the heater in both cases was the same (0.2 mm). The measurement data for the nanoparticle concentration of 0.05% are presented in Fig. 6, which shows the dependence of relative critical heat flux ðq=qw Þ on nanoparticle size. As is obvious, for both nanofluids the CHF increases with increasing nanoparticle diameter. At first glance, it seems paradoxical because with decreasing nanoparticle diameter, their number density in the fluid increases, and hence increases the number of nucleation centers. This seemingly should increase the critical heat flux with decreasing nanoparticle diameter. However, the measurements show just the opposite. Critical heat flux increases with increasing the diameter of nanoparticles. This fact can be interpreted by assuming that particle deposition on the heater surface plays the key role in the enhancement of CHF in nanofluids. The greater the size of the deposited particles, the greater the layer in which they are deposited, and the formation of the layer with necessary thickness is faster. Data on electron microscopy shown in Fig. 7 indirectly confirm this idea. It is seen that at the same particle concentration and the same boiling process duration, the condition of the heater surface with particles of different sizes is different. It is obvious that the larger the particle size, the better the coating
Fig. 4. SEM images of the nickel-chromium wire surface at different concentrations of SiO2 nanoparticles with the size of 25 nm. From left to right, top to bottom: distilled water, nanofluids with particle concentrations 0.05 and 0.1%, respectively.
218
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 5. SEM images of the nickel-chromium wire surface at different magnification after boiling in water-based nanofluids with SiO2 nanoparticles (dh ¼ 0.2 mm, 4 ¼ 0.1%, dp ¼ 25 nm).
quality. For particles of smaller size, there are significant areas on the heater not covered by the deposits. For particles of 40 nm the size of these areas reduces, while the 100 nm particles fully cover the heater surface. Thus, in this paper, we demonstrate a direct correlation between the heater surface condition and the critical heat flux in boiling nanofluids with different size of nanoparticles. 3.3. The dependence of CHF on the heater size In this section, we investigated the influence of the heater diameter, which varied within the range 0.1e0.3 mm. The data obtained for nanofluids with silica particles with an average size of 25 nm are presented in Fig. 8a, which shows the dependence of the relative critical heat flux ðq=qw Þ on the heater diameter. Here ðq=qw Þ is the critical heat flux for each of the heaters in distilled water. As is obvious, the critical heat flux significantly increases with decreasing the heater diameter at a fixed size of nanoparticles.
Again, such behavior can be explained by assuming that the deposition of nanoparticles on the heater surface plays the leading role in the nanofluid boiling process. The surface area of the heater with a diameter of 0.1 mm is nine times less than that of the heater with a diameter of 0.3 mm. Consequently, at the same concentration of nanoparticles in the fluid, the required height of deposits on the surface of a smaller heater is formed much faster than on the surface a larger heater. The conclusion is confirmed by Fig. 9, which presents the data on electronic microscopy of the heaters with various diameters. We see that a wire with the diameter of 0.1 mm is fully covered with a thick layer of sediments. The deposits on the wire with the diameter of 0.2 mm are not continuous. They are interrupted by spots of uncovered clean surface. The same pattern was observed in boiling on the wire with the diameter of 0.3 mm, were deposits were also very poor (see Fig. 9). Thus, the existence of correlation between the critical heat flux and the size of the heater is another indirect confirmation of the key role played by the particle deposition in the boiling crisis of nanofluids. This correlation may provide a possible explanation for the significant differences in the experimental data obtained by different researchers. The dependence of CHF on the heater size at the concentration of 0.5% is shown in Fig. 8b. As we see, this dependence is linear. Fig. 8a shows that CHF on the heater with the diameter of 0.1 mm is almost four times higher than that in pure water. When increasing the boiling surface area while maintaining all nanofluid parameters unchanged, the critical heat flux decreases monotonically. Therefore, the use of nanofluids for enhancing the critical heat flux will be most efficient in micro-devices, such as, for example, two-phase microchannel heat exchangers.
3.4. The dependence of CHF on the boiling process duration
Fig. 6. The dependence of the relative critical heat flux on nanoparticles diameter at the concentration of 0.05%.
The dependence of critical heat flux on the heater dimensions suggests that there must be a dependence of CHF on boiling process duration. To check this assumption, we have conducted a series of experiments with varying boiling process duration. The time variation was realized by changing the burden step and the delay in
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
219
Fig. 7. SEM images of the nickel-chromium wire surface after boiling of water-based nanofluid with SiO2 nanoparticles (dh ¼ 0.2 mm, 4 ¼ 0.05%, the boiling process duration is 25 min) of various sizes. From left to right, top to bottom: 25 nm, 40 nm, 100 nm.
equilibrium. We think that the existence of such temporal dependence revealed in the experiments can be considered as direct confirmation of particle deposition key role in the boiling crisis of nanofluids. Data on electron microscopy shown in Fig. 11 confirm this assumption. 3.5. The dependence CHF on the nanoparticle material
Fig. 8. The dependence of relative CHF on concentration for different heater diameters, dp ¼ 25 nm (a) and on the heater diameter for 4 ¼ 0.5% (b).
heater power supply. At the first step, a similar study was conducted for pure water. No dependence of CHF on the boiling process duration was found. Fig. 10 represents the dependence of the critical heat flux on boiling time for nanofluid containing SiO2 particles with the diameter of 25 nm at a volume concentration of 0.5%. As is obvious, for nanofluids there is a clear dependence between the critical heat flux and the boiling process duration. With increasing boiling duration, CHF increases rapidly and then reaches a certain steady state level. The dependence of the critical heat flux is well described by the correlation q=qw ¼ 1 þ at b , where a ¼ 0.734 and b ¼ 0.2. It is easy to explain such behavior based on the hypothesis that CHF is influenced by particle deposition on the heater. The growth and formation of deposits on the heater surface is not an instant process. It takes a certain time, after which the height of surface micro roughness ceases to change due to some state of dynamic
Another important factor is the dependence of critical heat flux on the nanoparticle material, which was previously observed by many researchers [12e25]. The dependence of CHF on the nanoparticle material and concentration is shown in Fig. 3. However, we have revealed that the CHF is strongly dependent on particle size, and therefore it is more correct to analyze the dependence on particle material at fixed particle size. This dependence is shown in Fig. 6. Here it is obvious that the critical heat flux during boiling of nanofluid with alumina particles is significantly lower (almost twice) than that with silica particles at the same particle size. The microstructure of the heater surface after boiling in these nanofluids is shown in Fig. 12a and b. It is evident that both the structure and thickness of deposits for various particles is different. For comparison, Fig. 12ced shows SEM images of the surface for other nanoparticles. As is obvious, the surface topography of deposits is different for various materials. For alumina particles, the deposits are formed in the shape of a long horizontal grooves with the width of 3e4 mm, while iron oxide particles form a very dense deposits with the size of roughness not exceeding 1 mm. The diamond particles are deposited in the form of flakes about 1e2 mm thick with the size of 30e40 mm, and the particles of silicon oxide cover the heater by several layers, in which holes with the sizes of 5e10 mm are visible (see Fig. 13). As is seen from Figs. 3 and 6, the structure of relief from nanoparticle deposits affects the critical heat flux. Under other conditions being equal (concentration, size, boiling process duration), the heaters having different nanoparticle deposit reliefs
220
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 9. The deposition of SiO2 nanoparticles with the size of 25 nm from the nanofluid on the surface of wires with different diameters of heater during 2 min. From left to right, top to bottom: 0.1 mm, 0.2 mm, 0.3 mm.
4. Conclusions
Fig. 10. The dependence of relative heat flux on the boiling process duration in the water-based nanofluid with SiO2 particles (4 ¼ 0.1%, dh ¼ 0.2 mm).
have different values of CHF. Of course, we cannot assert that the deposit relief has a direct impact on CHF. Most likely, the surface relief and the particle material affect the wettability of the heater surface. Many researchers [9e29] have indicated that the improvement of wettability is considered generally as the main CHF growth factor in boiling nanofluids. Different growth rates and deposit surface reliefs are apparently associated with the material properties of particles and the heater surface. Different nanoparticle materials have different adhesion with each other and with the heater material. Obviously, one must expect the effect of heater material on CHF. In addition, the wettability of particle material should essentially affect the deposits formation and its growth rate.
In this paper the experimental study of saturated boiling on a cylindrical heater has been carried out in various nanofluids. The studied nanofluids were prepared based on distilled water and nanoparticles of silicon, aluminum, and iron oxides as well as diamond. The volumetric concentration of the nanoparticles varied from 0.05 to 1%. The nanoparticle diameter ranged from 5 to 100 nm. The heater diameter was changed from 0.1 to 0.3 mm. The conducted experiments have shown that the use of nanofluids allows increasing significantly the critical heat flux in pool boiling even at very small nanoparticle concentrations (less than 1%). Even at the volume concentration of particles equal to a quarter percent, the critical heat flux increases by more than 50% and continues growing with a further increase in nanoparticle concentration. It is revealed that at high concentrations of nanoparticles, the growth rate of CHF slows down and it reaches a constant value. With the help of electron microscopy, it was shown that such behavior can be explained by the stabilization of deposits size on the heater surface. It is established that the critical heat flux in nanofluids depends on the nanoparticle size. Testing two different nanoparticle materials showed that the critical heat flux increases with increasing nanoparticle size. This fact can be explained by assuming that the particle deposition on the heater surface plays the key role in the CHF enhancement. The larger the size of deposited particles, the larger the scale of resulting roughness on the heater surface that facilitates the formation of deposit thickness necessary for enhanced boiling. This assumption is confirmed by the electron microscopy data. Thus, in this paper, we demonstrate a direct relationship between the heater surface condition and the critical heat flux during saturated pool boiling of nanofluids with different nanoparticle size. Besides, we have demonstrated the dependence
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
221
Fig. 11. SEM images of the nickel-chromium wire surface after boiling in water and in water-based nanofluids with SiO2 particles at different boiling process duration (dh ¼ 0.2 mm, 4 ¼ 0.1%, dp ¼ 25 nm; 500-fold multiplication).
Fig. 12. SEM images of nickel-chromium heater surface after boiling in different nanofluids (dh ¼ 0.2 mm, 4 ¼ 0.05%). From left to right, top to bottom: Al2O3 e 100 nm, SiO2 -100 nm, diamond e 5 nm, Fe3O4 e 95 nm.
222
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 13. SEM images of nickel-chromium heater surface after boiling in different nanofluids (dh ¼ 0.2 mm, 4 ¼ 0.05%). From left to right, top to bottom: Al2O3 e 100 nm, SiO2 e 100 nm, diamond e 5 nm, Fe3O4 e 95 nm.
of CHF on the particle material. Various nanofluids have different values of CHF at other conditions being equal (concentration, size, and boiling process duration). Different nanoparticle materials result in different thickness and surface relief of deposits on the heater surface. It is shown that with decreasing the heater diameter, the relative critical heat flux in nanofluids increases significantly. Such behavior can be explained by assuming that the deposition of nanoparticles on the heater surface plays the leading role in the nanofluid boiling process. At the same nanoparticle concentration in the fluid, the desired height of deposits on the surface of a smaller heater is formed much faster than that on the surface of a larger heater. It is evidenced by the data of electronic microscopy. The effect of the nanofluid boiling process time on the critical heat flux has been investigated as well. The conducted experiments revealed the existence of a certain correlation. With increasing boiling process time, the critical heat flux increases rapidly and then reaches some steady-state level. Thus, experimentally evidenced relationship between CHF and the concentration, particle size and material, as well as heater diameter and the boiling process duration can be considered as a direct confirmation of particle deposition key role in the nanofluids boiling crisis. Acknowledgment The work is performed at partial financial support of the Russian Science Foundation (Project No. 14-19-00312), SEM study is performed at partial financial support of the Krasnoyarsk Regional Fund for Support of Scientific and Scientific-Technical Activities (Contract N 16-48-243042\16). References [1] Kutateladze SS. Heat transfer in condensation and boiling. AEC-TR-3770. 1952.
[2] Rosenhow W, Griffith P. Correlation of maximum heat flux data for boiling of saturated fluids. Chem Eng Prog Symp Ser 1955;52(18):47e9. [3] Zuber N. Hydrodynamic aspects of boiling heat transfer. AEC Rep. AECU-4439. 1959. [4] Katto Y, Yokoya S. Principal mechanism of boiling crisis in pool boiling. Int J Heat Mass Transf 1968;11:993e6. http://dx.doi.org/10.1016/0017-9310(68) 90005-7. [5] You SM, Kim JH, Kim KH. Effect of nanoparticles on critical heat flux of water in pool boiling heat transfer. Appl Phys Lett 2003;83:3374e6. http://dx.doi. org/10.1063/1.1619206. [6] Das SK, Putra N, Roetzel W. Pool boiling of nanofluids on horizontal narrow tubes. Int J Multiph Flow 2003;29(8):1237e47. http://dx.doi.org/10.1016/ S0301-9322(03)00105-8. [7] Tsai CY, Chien HT, Ding PP, Chan B, Luh TY, Chen PH. Effect of structural character of gold nanoparticles in nanofluid on heat pipe thermal performance. Mater. Lett 2003;58:1461e5. http://dx.doi.org/10.1016/j.matlet.2003. 10.009. [8] Bang IC, Chang SH. Boiling heat transfer performance and phenomena of Al2O3ewater nanofluids from a plain surface in a pool. Int J Heat Mass Transf 2005;48(12):2407e19. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2004. 12.047. [9] Vassallo P, Kumar R, D'Amico S. Pool boiling heat transfer experiments in silicaewater nano-fluids. Int J Heat Mass Transf 2004;47(2):407e11. http://dx. doi.org/10.1016/S0017-9310(03)00361-2. [10] Tu JP, Dinh N, Theofanous T. An experimental study of nanofluid boiling heat transfer. In: Proceedings of the 6th international symposium on heat transfer; 2004. Beijing, China. [11] Witharana S. Boiling of refrigerants on enhanced surfaces and boiling of nanofluids. Ph.D. thesis. Stockholm, Sweden: Royal Institute of Technology; 2003. [12] Kim SJ, Bang IC, Buongiorno J, Hu LW. Effects of nanoparticle deposition on surface wettability influencing boiling heat transfer in nanofluids. App Phys Lett 2006;89(15):153107. http://dx.doi.org/10.1063/1.2360892. [13] Kim SJ, Bang IC, Buongiorno J, Hu LW. Surface wettability change during pool boiling of nanofluids and its effect on critical heat flux. Int J Heat Mass Transf 2007;50(19e20):4105e16. http://dx.doi.org/10.1016/j.ijheatmasstransfer. 2007.02.002. [14] Taylor RA, Phelan PE. Pool boiling of nanofluids: comprehensive review of existing data and limited new data. Int J Heat Mass Transf 2009;52(23e24): 5339e47. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2009.06.040. [15] Jo B, Jeon PS, Yoo J, Kim HJ. Wide range parametric study for the pool boiling of nanofluids with a circular plate heater. J Vis 2009;12(1):37e46. http:// dx.doi.org/10.1007/BF03181941. [16] Hyungdae K. Enhancement of critical heat flux in nucleate boiling of nanofluids: a state-of-art review. Nanoscale Res Lett 2011;6:415e20. http:// dx.doi.org/10.1186/1556-276X-6-415.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223 [17] Kim ES, Jung J-Y, Kang YT. The effect of surface area on pool boiling heat transfer coefficient and CHF of Al2O3/water nanofluids. J Mech Sci Technol 2013;27(10):3177e82. http://dx.doi.org/10.1007/s12206-013-0839-7. [18] Lu M-C, Chen R, Srinivasan V, Carey VP, Majumdar A. Critical heat flux of pool boiling on Si nanowire array-coated surfaces. Int J Heat Mass Transf 2011;54(25e26):5359e67. http://dx.doi.org/10.1016/j.ijheatmasstransfer. 2011.08.007. [19] Sakashita H. Pressure effect on CHF enhancement in pool boiling of nanofluids. J Nucl Sci Technol 2015:797e802. http://dx.doi.org/10.1080/00223131. 2015.1072482. [20] Moreno GJ, Oldenburg S, You SM, Kim JH. Pool boiling heat transfer of alumina-water, zinc oxide-water and alumina-water ethylene glycol nanofluids. Proceedings of the Int. Conf. HT2005 2005;2:625e32. http://dx.doi.org/ 10.1115/HT2005-72375. San Francisco, California, USA. [21] Kwark SM, Amaya M, Kumar R, Moreno G, You SM. Effects of pressure, orientation, and heater size on pool boiling of water with nanocoated heaters. Int J Heat Mass Transf 2010;53(23e24):5199e208. http://dx.doi.org/10.1016/ j.ijheatmasstransfer.2010.07.040. [22] Pryazhnikov MI, Minakov AV, Rudyak VYa. The influence of the size and material of nanoparticles and the heater size on the critical heat flux density in boiling nanofluids. Tech Phys Lett 2015;41(9):891e3. http://dx.doi.org/ 10.1134/S1063785015090278. [23] Minakov AV, Lobasov AS, Rudyak VYa, Guzei DV, Pryazhnikov MI. Measuring
[24]
[25]
[26]
[27]
[28]
[29]
223
the heat-transfer coefficient of nanofluid based on copper oxide in a cylindrical channel. Tech Phys Lett 2014;40(3):203e6. http://dx.doi.org/10.1134/ S1063785014030067. Kuzma-Kichta YuA, Lavrikov AV, Shustov MV, Chursin PS, Chistyakova AV. Boiling characteristics at nanoparticle coated surface. In: Proceedings of the ECI 8th international conference on boiling and condensation heat transfer; 2012. Lausanne, Switzerland. Kuzma-Kichta YuA, Lavrikov AV, Shustov MV, Tsurikov D, Zhukov V. Boiling heat transfer enhancement on macro-, micro- and nano-scales. J Nucl Eng Technol 2012;2(2). Li Xiangdong, Cheung Sherman Chi Pok, Tu Jiyuan. Nucleate boiling of dilute nanofluids: mechanism of exploring and modeling. Int J Therm Sci 2014;84: 323e34. http://dx.doi.org/10.1016/j.ijthermalsci.2014.05.021. Li X, Li K, Tu J, Buongiorno J. On two-fluid modeling of nucleate boiling of dilute nanofluids. Int J Heat Mass Transf 2014;69:443e50. http://dx.doi.org/ 10.1016/j.ijheatmasstransfer.2013.10.037. Kamatchi R, Venkatachalapath S. Parametric study of pool boiling heat transfer with nanofluids for the enhancement of critical heat flux: a review. Int J Therm Sci 2015;87:228e40. http://dx.doi.org/10.1016/j.ijthermalsci. 2014.09.001. Bi J, Vafai K, Christopher DM. Heat transfer characteristics and CHF prediction in nanofluid boiling. Int J Heat Mass Transf 2015;80:256e65. http://dx.doi. org/10.1016/j.ijheatmasstransfer.2014.09.019.
Contents lists available at ScienceDirect
International Journal of Thermal Sciences journal homepage: www.elsevier.com/locate/ijts
The experimental study of nanofluids boiling crisis on cylindrical heaters A.V. Minakov a, b, c, *, M.I. Pryazhnikov a, b, c, D.V. Guzei a, c, G.M. Zeer a, V. Ya. Rudyak c a
Siberian Federal University, Krasnoyarsk, Russia Kutateladze Institute of Thermophysics, SB RAS, Novosibirsk, Russia c Novosibirsk State University of Architecture and Civil Engineering, Russia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 April 2016 Received in revised form 26 January 2017 Accepted 27 February 2017
The paper deals with the experimental study of nanofluids boiling on cylindrical heater. The studied nanofluids were prepared on the basis of distilled water and nanoparticles of silicon, aluminum and iron oxides as well as diamond. The volume concentration of nanoparticles varied from 0.05 to 1%. The nanoparticles diameter ranged from 10 to 100 nm, the diameter of the heater was varied from 0.1 to 0.3 mm. It is revealed that the use of nanofluids provides a several-fold increase of the critical heat flux. However, critical heat flux in nanofluids depends on the material and size of nanoparticles as well as on the diameter of used cylindrical heater. It is shown that the critical heat flux increases with increasing the nanoparticles size, while it decreases with increasing the heater diameter. It was revealed for the first time that the critical heat flux depends on the boiling process duration. © 2017 Elsevier Masson SAS. All rights reserved.
Keywords: Nanofluid Nanoparticles Boiling crisis Critical heat flux (CHF) Deposition of nanoparticles Heater surface microstructure
1. Introduction Boiling, as one of the most effective and efficient ways of heat transfer, is used in various engineering applications. Therefore, the enhancement of the critical heat flux in boiling has been the subject of numerous studies carried out in the last century [1e4]. An intensive research on the possible use of nanofluids in boiling-related applications has started just recently. Apparently, works [5,6] should be considered as the first investigations of nanofluids boiling heat transfer and crisis. The first work dealt with the boiling of water-based silicon dioxide and aluminum oxide nanofluids on a square heater with a characteristic size of 10 mm. The authors observed a sharp increase in the critical heat flux (CHF) caused by the presence of nanoparticles. At the pressure of about 0.198 bar, the increase in CHF reached more than 200% when using water-based Al2O3 nanofluid with nanoparticle concentration of 5$104 wt%. It was also noted that the supplement of nanoparticles to water increases the size of bubbles, though reduces the frequency of their detachment. It is unclear, however, how these changes are related to the observed increase in CHF.
* Corresponding author. Siberian Federal University, Krasnoyarsk, Russia. E-mail address: [email protected] (A.V. Minakov). http://dx.doi.org/10.1016/j.ijthermalsci.2017.02.019 1290-0729/© 2017 Elsevier Masson SAS. All rights reserved.
In Ref. [6], a nucleate pool boiling was studied on the surface of cylindrical heating element 20 mm in diameter in the water-based alumina nanofluid. Contrary to work [5], it was shown that the presence of nanoparticles degrades the boiling characteristics (heat transfer coefficient), and this deterioration increases with increasing nanoparticle concentration. A similar phenomenon was observed in a later study [7], where smaller heaters with an outer diameter ranged from 4.5 to 6 mm were used. The authors explained the deterioration in heat transfer by the changes of heaters surface characteristics. They claim that during boiling of nanofluids, the surface became smoother due to the deposition of nanoparticles at the nucleation sites. The higher the concentration, the smoother the surface, which in turn leads to a greater reduction of heat transfer coefficient. This explanation is not consistent with observations [8], concerned with the study of water-based alumina nanofluids boiling on a square surface with the characteristic size of 100 mm at high heat fluxes. It was revealed that after boiling, the surface roughness increased with increasing the nanoparticle concentration. Vassallo et al. [9] studied boiling of water-based silicon dioxide nanofluid on the horizontal nickel-chromium wire with a diameter of 0.4 mm at atmospheric pressure. Similar to work [5], a two-fold increase in CHF was observed.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Tu et al. [10] conducted boiling experiments in water-based alumina nanofluid in a relatively large volume with rectangular heating surface of 26 40 mm2 at atmospheric pressure. A significant increase in both heat transfer and CHF was obtained in the nucleate boiling mode. For example, compared to pure water boiling, an increase in heat transfer coefficient by more than 60% was recorded while investigating highly diluted aqueous nanofluids. A significant reduction in the maximum wall temperature of the heater as well as a four-fold increase in the number of active bubble nucleation sites was found in comparison with pure water boiling. The boiling heat transfer in water-based nanofluid with gold nanoparticles was studied in Ref. [11]. A significant increase in heat transfer coefficient was found during boiling at atmospheric pressure on the heater with the diameter of 100 mm. In particular, the authors managed to increase the heat transfer coefficient of nanofluid containing 0.001 wt % of gold nanoparticles by 21% at the heat flux of 45 kW/m2. The increase in boiling heat transfer coefficient was also accompanied by the growth of critical heat flux with increasing nanoparticle concentration. Most researchers observe the deposition of nanoparticles on the heater surface in the course of boiling. Apparently [12], was the first work, where the increase in CHF was explained by the presence of nanoparticles depositions (see also [13]). It was shown that the wettability of heater surface covered by deposited nanoparticles was much higher than that of a pure surface. The key role of nanoparticles deposition on the heater surface in the boiling crisis and the effect of surface wettability on the CHF were also observed in the later works [14e28]. The effect of nanoparticle concentration on CHF has been studied well enough to date. It is revealed that in most cases, increasing concentration of nanoparticles increases CHF, though in some cases there was a decrease in CHF at high concentrations of nanoparticles [15]. In addition, the effect of heater surface area on the CHF was observed in Refs. [17,18]. It is shown in Ref. [18] that increasing the size of the heater slows down the enhancement of CHF in nanofluids. The effect of pressure on CHF was studied in Ref. [19], where it was revealed that the gain in CHF the value due to use of nanofluids instead of pure fluids decreases with the increase of pressure. The data on the effect of particle size on CHF are rather contradictory. Thus, Moreno et al. [20] claimed that CHF does not depend on the size of alumina nanoparticles within the particle size range from 69 to 346 nm. Similar conclusions were made in Ref. [21]. At the same time, it was found in Ref. [15] that CHF decreases with increasing particle size in nanofluids containing silver particles. On the contrary, in Refs. [22,23] it was shown that CHF increases with increasing particles size. In many studies, it was observed that the effect of increased CHF was retained even in boiling of pure fluids on the heaters covered by nanoparticles [18,24,25]. Among the recent works, we should note theoretical papers and surveys [26e29], which are dealing with detail consideration of various mechanisms leading to enhancement of CHF in boiling nanofluids. However, the decisive conclusions on these mechanisms have not yet been made. Thus, despite a large number of experimental data, they are quite contradictory and do not allow comprehending clearly the mechanisms of CHF variations in the boiling process of nanofluids. In order to remove at least some of these contradictions, this paper provides with the systematic experimental study of saturated pool boiling of nanofluids on a cylindrical heater. Our main objective is to understand the dependence of CHF on the nanoparticle size and material, heat diameter, and nanoparticles deposition on its surface.
215
2. Description of the experimental setup The scheme of experimental setup is presented in Fig. 1. The studied fluid in amount of 300 ml was placed in a high sealed glass flask (4) with the diameter of 8 cm. A nickel-chromium wire heater with the length L ¼ 34 mm and the diameter dh ¼ 0.2 mm was emerged into the flask filled with fluid. The wire was fixed by copper bus leads with the cross section of 10 2 mm to supply voltage. The heater was energized by means of GPS-6030D power source. The applied voltage was recorded using a GDМ-78261 voltmeter (“V” in Fig. 1). Load amperage in the circuit was measured by the APPAe109N ammeter (“A” in Fig. 1). The flask with the test fluid was sealed, so that the condensate formed in the upper part of the flask dripped back into the flask, maintaining saturation conditions in the working chamber. The fluid temperature in the flask was controlled by means of a chromel-copel thermocouple, which was connected to the TRM-200 temperature meter (5). The tail end of the thermocouple was located at the same level with the nickelchromium heater at a distance of 2 cm from it. The flask with the test fluid was placed in a water bath (3) with the constant temperature, which was about 0.5 C below the boiling point and maintained by means of electric heater (2). Thus, in this paper we investigate the boiling close to saturation conditions. The temperature control in a water bath was conducted using another chromel-copel thermocouple connected to the TRM200 m. After placing the flask into a water bath, the temperatures in the flask and water bath equalized with time. Then the nickelchromium heater was energized, and the heat flux density was measured. The programmable current power supply allowed increasing the heater voltage with a fixed predetermined step. Thus, it was possible to control the heat flux growth rate and, thereby, to fix the onset of boiling crisis. The boiling heat flux density on the heater was determined by the ratio: q ¼ Q =S ¼ IU=S, where S ¼ PdL is the lateral surface area of the heater between the current-carrying wires, Q is the heat flux released by the heater, I is the electric current in the heater circuit,
Fig. 1. The scheme of experimental setup.
216
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
U is the voltage drop in the heater. Heat generated at the lead wires was negligible. The total error in determining the heat flux density was about 2%. Studying the boiling processes requires also the knowledge of heater surface temperature. For the nickel-chromium, the dependence of resistance R on temperature within the investigated temperature range is close to linear R ¼ R0 ½1 þ AðT T0 Þ. Thus, by measuring the resistance of the heater and knowing the temperature coefficient of resistance A, we can find the average heater temperature: T ¼ T0 þ ðR=R0 1Þ=A. Here R0 is the heater resistance at the temperature T0 . To determine the temperature coefficient of nickel-chromium electrical resistance, we have performed special series of resistance measurements of the used nickelchromium wire. It was found that A ¼ 0.00012 1/K. The total error in determining the heater temperature was about 3%. The nanofluids were prepared based on distilled water and nanoparticles of aluminum, silicon, iron (III) oxides as well as nanodiamonds. The volume concentration of particles 4 ranged from 0.05 to 1%. For preparation of nanofluids we used the standard two-step process. After adding the necessary quantity of nanopowder to water, the obtained nanofluid was first thoroughly mixed mechanically. After that, it was placed in an ultrasonic disperser for half an hour to destruct the particles conglomerates. No surfactants were added to the nanofluid. The diameter of the used nanoparticles dp ranged from 10 to 100 nm. The following particle sizes were used: SiO2 (25, 40, 100 nm), Al2O3 (10, 50, 100 nm), Fe2O3 (95 nm), and the diamond particles (5 nm). The first three types of nanopowder were received from “Plazmoterm” company, while nanodiamonds were acquired from FGUP Altay. Measurement of nanoparticles distribution in the fluid in terms of their size was carried out by means of CPS Disk Centrifuge DC24000. Besides measuring the critical heat flux we have investigated also the heaters surface structure after boiling. It was done by using transmission electron microscope JEM-2100 (JEOL, Japan) equipped with an energy dispersive spectrometer Oxford Inca x-sight. 3. Measurement results
Fig. 2. Boiling curves in distilled water and nanofluids with different concentrations of SiO2 particles.
dependence of the relative critical heat flux q=qw (where qw is CHF for water) on the nanoparticle volume concentration for three water-based nanofluids with particles of silicon dioxide (25 nm), iron oxide (95 nm), and diamond (5 nm). The values of critical heat flux presented here were obtained by averaging the data of five independent measurements. The CHF for water was equal to qw ¼ 1:32 MW/m2, which is consistent with the data [12,13]. The analysis of the boiling heat flux behavior in nanofluids shows that at low particle concentrations it increases drastically. However, the growth rate decreases with increasing particle concentration. We can expect that starting from a certain particle concentration CHF will no longer continue to increase. The dependence of the relative critical thermal load q=qw on the particle concentration shown in Fig. 3 significantly differs for the three nanofluids considered. However, in all cases this dependence can be described by the equation
3.1. The dependence of CHF on the nanoparticle concentration The first series of experiments was carried out to determine the dependence of CHF on the nanoparticle concentration. Typical results are presented in Fig. 2 for the water-based nanofluid with SiO2 particles at an average particle size dp of 25 nm and the heater diameter dh of 0.2 mm. The figure shows the dependence of heat flux density on the superheat DT ¼ Tw Tf (where Tw is the heater surface temperature and Tf is the fluid temperature). Even at the nanoparticles concentration of 0.05%, we observed almost two-fold increase in CHF for nanofluid in comparison with that for pure water. This difference in CHF increases with increasing concentration of nanoparticles. At the nanoparticle concentration of 0.5% it becomes almost three-fold. After reaching the critical heat load there was a transition from nucleate boiling to film boiling mode. At that, the heater either burned out instantly or, in very rare cases, continued generating heat in film boiling mode. This a behavior is shown in Fig. 2 for water (see the growth of CHF, which starts at the temperatures above 700 C). In this case, we failed to reach film boiling mode in nanofluids, since the heater was almost always burned out because the thermal load corresponding to CHF in nanofluids during the transition to film boiling mode was 2e3 times higher than that in pure water. The increase in CHF with increasing volume concentration of particles was observed for all nanofluids. Fig. 3 shows the
q=qw ¼ 1 þ a4b
(1)
The constants in equation (1) in general depend on several
Fig. 3. The dependence of the relative critical heat flux on the nanoparticle volume concentration.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
parameters, although for the studied nanofluids they can be taken as follows: a ¼ 2.169, b ¼ 0.1903 for SiO2 particles; a ¼ 1.125, b ¼ 0.4117 for diamond particles; and a ¼ 0.8595, b ¼ 0.3292 for particles of iron oxide Fe3O4. In the majority of works concerned with the nanofluids boiling crisis it is noted that the main reason for increasing of CHF is the nanoparticles deposition on the heater surface. The microstructure of the 0.2 mm heater surface sample and nanofluid coating containing SiO2 particles with an average size of 25 nm is shown in Fig. 4 for different particle concentrations. In all the cases, the boiling process duration was 25 min. It can be seen that the heater surface is covered by fairly dense deposits. With increasing nanoparticle concentration in the fluid at other conditions being equal, the deposits thickness and the proportion of deposited area increase. Thus, the critical heat flux is directly related to deposits of nanoparticles on the heater surface. The fact that these deposits are composed of individual nanoparticles is clearly seen from Fig. 5f at a maximum magnification of 30,000 times. The electron microscopy has shown that at high nanoparticle concentrations, the thickness of deposits reaches about 5e10 mm and further slightly varies with time and with increasing nanoparticle concentration. The analysis of the surface microstructure allows us to conclude that starting from a certain time moment, there exists a dynamic equilibrium between the number of particles deposited on the heater and those washed back into the bulk of fluid. Apparently, it explains the noted slowdown in the growth of CHF and its ultimate saturation caused by increase in particles concentration. It should be noted that the deposits on the heater surface are not solid. The bulk of formed deposits consist of developed microchannels network with a thickness of 5e10 mm and a length of 10e50 mm.
217
3.2. The dependence of CHF on the nanoparticle size The comparison presented in Fig. 3 shows a significant dependence of the relative CHF on nanoparticle volume concentration. However, the three curves shown there are obtained for nanofluids with different particles (in particular, they have different sizes). So far, there are no indisputable data on the dependence of critical heat flux on the nanoparticle size. To ascertain the nature of this dependency we have carried out special series of measurements. Two water-based nanofluids with silica and alumina particles of different size ranged from 10 to 100 nm were considered. The diameter of the heater in both cases was the same (0.2 mm). The measurement data for the nanoparticle concentration of 0.05% are presented in Fig. 6, which shows the dependence of relative critical heat flux ðq=qw Þ on nanoparticle size. As is obvious, for both nanofluids the CHF increases with increasing nanoparticle diameter. At first glance, it seems paradoxical because with decreasing nanoparticle diameter, their number density in the fluid increases, and hence increases the number of nucleation centers. This seemingly should increase the critical heat flux with decreasing nanoparticle diameter. However, the measurements show just the opposite. Critical heat flux increases with increasing the diameter of nanoparticles. This fact can be interpreted by assuming that particle deposition on the heater surface plays the key role in the enhancement of CHF in nanofluids. The greater the size of the deposited particles, the greater the layer in which they are deposited, and the formation of the layer with necessary thickness is faster. Data on electron microscopy shown in Fig. 7 indirectly confirm this idea. It is seen that at the same particle concentration and the same boiling process duration, the condition of the heater surface with particles of different sizes is different. It is obvious that the larger the particle size, the better the coating
Fig. 4. SEM images of the nickel-chromium wire surface at different concentrations of SiO2 nanoparticles with the size of 25 nm. From left to right, top to bottom: distilled water, nanofluids with particle concentrations 0.05 and 0.1%, respectively.
218
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 5. SEM images of the nickel-chromium wire surface at different magnification after boiling in water-based nanofluids with SiO2 nanoparticles (dh ¼ 0.2 mm, 4 ¼ 0.1%, dp ¼ 25 nm).
quality. For particles of smaller size, there are significant areas on the heater not covered by the deposits. For particles of 40 nm the size of these areas reduces, while the 100 nm particles fully cover the heater surface. Thus, in this paper, we demonstrate a direct correlation between the heater surface condition and the critical heat flux in boiling nanofluids with different size of nanoparticles. 3.3. The dependence of CHF on the heater size In this section, we investigated the influence of the heater diameter, which varied within the range 0.1e0.3 mm. The data obtained for nanofluids with silica particles with an average size of 25 nm are presented in Fig. 8a, which shows the dependence of the relative critical heat flux ðq=qw Þ on the heater diameter. Here ðq=qw Þ is the critical heat flux for each of the heaters in distilled water. As is obvious, the critical heat flux significantly increases with decreasing the heater diameter at a fixed size of nanoparticles.
Again, such behavior can be explained by assuming that the deposition of nanoparticles on the heater surface plays the leading role in the nanofluid boiling process. The surface area of the heater with a diameter of 0.1 mm is nine times less than that of the heater with a diameter of 0.3 mm. Consequently, at the same concentration of nanoparticles in the fluid, the required height of deposits on the surface of a smaller heater is formed much faster than on the surface a larger heater. The conclusion is confirmed by Fig. 9, which presents the data on electronic microscopy of the heaters with various diameters. We see that a wire with the diameter of 0.1 mm is fully covered with a thick layer of sediments. The deposits on the wire with the diameter of 0.2 mm are not continuous. They are interrupted by spots of uncovered clean surface. The same pattern was observed in boiling on the wire with the diameter of 0.3 mm, were deposits were also very poor (see Fig. 9). Thus, the existence of correlation between the critical heat flux and the size of the heater is another indirect confirmation of the key role played by the particle deposition in the boiling crisis of nanofluids. This correlation may provide a possible explanation for the significant differences in the experimental data obtained by different researchers. The dependence of CHF on the heater size at the concentration of 0.5% is shown in Fig. 8b. As we see, this dependence is linear. Fig. 8a shows that CHF on the heater with the diameter of 0.1 mm is almost four times higher than that in pure water. When increasing the boiling surface area while maintaining all nanofluid parameters unchanged, the critical heat flux decreases monotonically. Therefore, the use of nanofluids for enhancing the critical heat flux will be most efficient in micro-devices, such as, for example, two-phase microchannel heat exchangers.
3.4. The dependence of CHF on the boiling process duration
Fig. 6. The dependence of the relative critical heat flux on nanoparticles diameter at the concentration of 0.05%.
The dependence of critical heat flux on the heater dimensions suggests that there must be a dependence of CHF on boiling process duration. To check this assumption, we have conducted a series of experiments with varying boiling process duration. The time variation was realized by changing the burden step and the delay in
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
219
Fig. 7. SEM images of the nickel-chromium wire surface after boiling of water-based nanofluid with SiO2 nanoparticles (dh ¼ 0.2 mm, 4 ¼ 0.05%, the boiling process duration is 25 min) of various sizes. From left to right, top to bottom: 25 nm, 40 nm, 100 nm.
equilibrium. We think that the existence of such temporal dependence revealed in the experiments can be considered as direct confirmation of particle deposition key role in the boiling crisis of nanofluids. Data on electron microscopy shown in Fig. 11 confirm this assumption. 3.5. The dependence CHF on the nanoparticle material
Fig. 8. The dependence of relative CHF on concentration for different heater diameters, dp ¼ 25 nm (a) and on the heater diameter for 4 ¼ 0.5% (b).
heater power supply. At the first step, a similar study was conducted for pure water. No dependence of CHF on the boiling process duration was found. Fig. 10 represents the dependence of the critical heat flux on boiling time for nanofluid containing SiO2 particles with the diameter of 25 nm at a volume concentration of 0.5%. As is obvious, for nanofluids there is a clear dependence between the critical heat flux and the boiling process duration. With increasing boiling duration, CHF increases rapidly and then reaches a certain steady state level. The dependence of the critical heat flux is well described by the correlation q=qw ¼ 1 þ at b , where a ¼ 0.734 and b ¼ 0.2. It is easy to explain such behavior based on the hypothesis that CHF is influenced by particle deposition on the heater. The growth and formation of deposits on the heater surface is not an instant process. It takes a certain time, after which the height of surface micro roughness ceases to change due to some state of dynamic
Another important factor is the dependence of critical heat flux on the nanoparticle material, which was previously observed by many researchers [12e25]. The dependence of CHF on the nanoparticle material and concentration is shown in Fig. 3. However, we have revealed that the CHF is strongly dependent on particle size, and therefore it is more correct to analyze the dependence on particle material at fixed particle size. This dependence is shown in Fig. 6. Here it is obvious that the critical heat flux during boiling of nanofluid with alumina particles is significantly lower (almost twice) than that with silica particles at the same particle size. The microstructure of the heater surface after boiling in these nanofluids is shown in Fig. 12a and b. It is evident that both the structure and thickness of deposits for various particles is different. For comparison, Fig. 12ced shows SEM images of the surface for other nanoparticles. As is obvious, the surface topography of deposits is different for various materials. For alumina particles, the deposits are formed in the shape of a long horizontal grooves with the width of 3e4 mm, while iron oxide particles form a very dense deposits with the size of roughness not exceeding 1 mm. The diamond particles are deposited in the form of flakes about 1e2 mm thick with the size of 30e40 mm, and the particles of silicon oxide cover the heater by several layers, in which holes with the sizes of 5e10 mm are visible (see Fig. 13). As is seen from Figs. 3 and 6, the structure of relief from nanoparticle deposits affects the critical heat flux. Under other conditions being equal (concentration, size, boiling process duration), the heaters having different nanoparticle deposit reliefs
220
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 9. The deposition of SiO2 nanoparticles with the size of 25 nm from the nanofluid on the surface of wires with different diameters of heater during 2 min. From left to right, top to bottom: 0.1 mm, 0.2 mm, 0.3 mm.
4. Conclusions
Fig. 10. The dependence of relative heat flux on the boiling process duration in the water-based nanofluid with SiO2 particles (4 ¼ 0.1%, dh ¼ 0.2 mm).
have different values of CHF. Of course, we cannot assert that the deposit relief has a direct impact on CHF. Most likely, the surface relief and the particle material affect the wettability of the heater surface. Many researchers [9e29] have indicated that the improvement of wettability is considered generally as the main CHF growth factor in boiling nanofluids. Different growth rates and deposit surface reliefs are apparently associated with the material properties of particles and the heater surface. Different nanoparticle materials have different adhesion with each other and with the heater material. Obviously, one must expect the effect of heater material on CHF. In addition, the wettability of particle material should essentially affect the deposits formation and its growth rate.
In this paper the experimental study of saturated boiling on a cylindrical heater has been carried out in various nanofluids. The studied nanofluids were prepared based on distilled water and nanoparticles of silicon, aluminum, and iron oxides as well as diamond. The volumetric concentration of the nanoparticles varied from 0.05 to 1%. The nanoparticle diameter ranged from 5 to 100 nm. The heater diameter was changed from 0.1 to 0.3 mm. The conducted experiments have shown that the use of nanofluids allows increasing significantly the critical heat flux in pool boiling even at very small nanoparticle concentrations (less than 1%). Even at the volume concentration of particles equal to a quarter percent, the critical heat flux increases by more than 50% and continues growing with a further increase in nanoparticle concentration. It is revealed that at high concentrations of nanoparticles, the growth rate of CHF slows down and it reaches a constant value. With the help of electron microscopy, it was shown that such behavior can be explained by the stabilization of deposits size on the heater surface. It is established that the critical heat flux in nanofluids depends on the nanoparticle size. Testing two different nanoparticle materials showed that the critical heat flux increases with increasing nanoparticle size. This fact can be explained by assuming that the particle deposition on the heater surface plays the key role in the CHF enhancement. The larger the size of deposited particles, the larger the scale of resulting roughness on the heater surface that facilitates the formation of deposit thickness necessary for enhanced boiling. This assumption is confirmed by the electron microscopy data. Thus, in this paper, we demonstrate a direct relationship between the heater surface condition and the critical heat flux during saturated pool boiling of nanofluids with different nanoparticle size. Besides, we have demonstrated the dependence
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
221
Fig. 11. SEM images of the nickel-chromium wire surface after boiling in water and in water-based nanofluids with SiO2 particles at different boiling process duration (dh ¼ 0.2 mm, 4 ¼ 0.1%, dp ¼ 25 nm; 500-fold multiplication).
Fig. 12. SEM images of nickel-chromium heater surface after boiling in different nanofluids (dh ¼ 0.2 mm, 4 ¼ 0.05%). From left to right, top to bottom: Al2O3 e 100 nm, SiO2 -100 nm, diamond e 5 nm, Fe3O4 e 95 nm.
222
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223
Fig. 13. SEM images of nickel-chromium heater surface after boiling in different nanofluids (dh ¼ 0.2 mm, 4 ¼ 0.05%). From left to right, top to bottom: Al2O3 e 100 nm, SiO2 e 100 nm, diamond e 5 nm, Fe3O4 e 95 nm.
of CHF on the particle material. Various nanofluids have different values of CHF at other conditions being equal (concentration, size, and boiling process duration). Different nanoparticle materials result in different thickness and surface relief of deposits on the heater surface. It is shown that with decreasing the heater diameter, the relative critical heat flux in nanofluids increases significantly. Such behavior can be explained by assuming that the deposition of nanoparticles on the heater surface plays the leading role in the nanofluid boiling process. At the same nanoparticle concentration in the fluid, the desired height of deposits on the surface of a smaller heater is formed much faster than that on the surface of a larger heater. It is evidenced by the data of electronic microscopy. The effect of the nanofluid boiling process time on the critical heat flux has been investigated as well. The conducted experiments revealed the existence of a certain correlation. With increasing boiling process time, the critical heat flux increases rapidly and then reaches some steady-state level. Thus, experimentally evidenced relationship between CHF and the concentration, particle size and material, as well as heater diameter and the boiling process duration can be considered as a direct confirmation of particle deposition key role in the nanofluids boiling crisis. Acknowledgment The work is performed at partial financial support of the Russian Science Foundation (Project No. 14-19-00312), SEM study is performed at partial financial support of the Krasnoyarsk Regional Fund for Support of Scientific and Scientific-Technical Activities (Contract N 16-48-243042\16). References [1] Kutateladze SS. Heat transfer in condensation and boiling. AEC-TR-3770. 1952.
[2] Rosenhow W, Griffith P. Correlation of maximum heat flux data for boiling of saturated fluids. Chem Eng Prog Symp Ser 1955;52(18):47e9. [3] Zuber N. Hydrodynamic aspects of boiling heat transfer. AEC Rep. AECU-4439. 1959. [4] Katto Y, Yokoya S. Principal mechanism of boiling crisis in pool boiling. Int J Heat Mass Transf 1968;11:993e6. http://dx.doi.org/10.1016/0017-9310(68) 90005-7. [5] You SM, Kim JH, Kim KH. Effect of nanoparticles on critical heat flux of water in pool boiling heat transfer. Appl Phys Lett 2003;83:3374e6. http://dx.doi. org/10.1063/1.1619206. [6] Das SK, Putra N, Roetzel W. Pool boiling of nanofluids on horizontal narrow tubes. Int J Multiph Flow 2003;29(8):1237e47. http://dx.doi.org/10.1016/ S0301-9322(03)00105-8. [7] Tsai CY, Chien HT, Ding PP, Chan B, Luh TY, Chen PH. Effect of structural character of gold nanoparticles in nanofluid on heat pipe thermal performance. Mater. Lett 2003;58:1461e5. http://dx.doi.org/10.1016/j.matlet.2003. 10.009. [8] Bang IC, Chang SH. Boiling heat transfer performance and phenomena of Al2O3ewater nanofluids from a plain surface in a pool. Int J Heat Mass Transf 2005;48(12):2407e19. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2004. 12.047. [9] Vassallo P, Kumar R, D'Amico S. Pool boiling heat transfer experiments in silicaewater nano-fluids. Int J Heat Mass Transf 2004;47(2):407e11. http://dx. doi.org/10.1016/S0017-9310(03)00361-2. [10] Tu JP, Dinh N, Theofanous T. An experimental study of nanofluid boiling heat transfer. In: Proceedings of the 6th international symposium on heat transfer; 2004. Beijing, China. [11] Witharana S. Boiling of refrigerants on enhanced surfaces and boiling of nanofluids. Ph.D. thesis. Stockholm, Sweden: Royal Institute of Technology; 2003. [12] Kim SJ, Bang IC, Buongiorno J, Hu LW. Effects of nanoparticle deposition on surface wettability influencing boiling heat transfer in nanofluids. App Phys Lett 2006;89(15):153107. http://dx.doi.org/10.1063/1.2360892. [13] Kim SJ, Bang IC, Buongiorno J, Hu LW. Surface wettability change during pool boiling of nanofluids and its effect on critical heat flux. Int J Heat Mass Transf 2007;50(19e20):4105e16. http://dx.doi.org/10.1016/j.ijheatmasstransfer. 2007.02.002. [14] Taylor RA, Phelan PE. Pool boiling of nanofluids: comprehensive review of existing data and limited new data. Int J Heat Mass Transf 2009;52(23e24): 5339e47. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2009.06.040. [15] Jo B, Jeon PS, Yoo J, Kim HJ. Wide range parametric study for the pool boiling of nanofluids with a circular plate heater. J Vis 2009;12(1):37e46. http:// dx.doi.org/10.1007/BF03181941. [16] Hyungdae K. Enhancement of critical heat flux in nucleate boiling of nanofluids: a state-of-art review. Nanoscale Res Lett 2011;6:415e20. http:// dx.doi.org/10.1186/1556-276X-6-415.
A.V. Minakov et al. / International Journal of Thermal Sciences 116 (2017) 214e223 [17] Kim ES, Jung J-Y, Kang YT. The effect of surface area on pool boiling heat transfer coefficient and CHF of Al2O3/water nanofluids. J Mech Sci Technol 2013;27(10):3177e82. http://dx.doi.org/10.1007/s12206-013-0839-7. [18] Lu M-C, Chen R, Srinivasan V, Carey VP, Majumdar A. Critical heat flux of pool boiling on Si nanowire array-coated surfaces. Int J Heat Mass Transf 2011;54(25e26):5359e67. http://dx.doi.org/10.1016/j.ijheatmasstransfer. 2011.08.007. [19] Sakashita H. Pressure effect on CHF enhancement in pool boiling of nanofluids. J Nucl Sci Technol 2015:797e802. http://dx.doi.org/10.1080/00223131. 2015.1072482. [20] Moreno GJ, Oldenburg S, You SM, Kim JH. Pool boiling heat transfer of alumina-water, zinc oxide-water and alumina-water ethylene glycol nanofluids. Proceedings of the Int. Conf. HT2005 2005;2:625e32. http://dx.doi.org/ 10.1115/HT2005-72375. San Francisco, California, USA. [21] Kwark SM, Amaya M, Kumar R, Moreno G, You SM. Effects of pressure, orientation, and heater size on pool boiling of water with nanocoated heaters. Int J Heat Mass Transf 2010;53(23e24):5199e208. http://dx.doi.org/10.1016/ j.ijheatmasstransfer.2010.07.040. [22] Pryazhnikov MI, Minakov AV, Rudyak VYa. The influence of the size and material of nanoparticles and the heater size on the critical heat flux density in boiling nanofluids. Tech Phys Lett 2015;41(9):891e3. http://dx.doi.org/ 10.1134/S1063785015090278. [23] Minakov AV, Lobasov AS, Rudyak VYa, Guzei DV, Pryazhnikov MI. Measuring
[24]
[25]
[26]
[27]
[28]
[29]
223
the heat-transfer coefficient of nanofluid based on copper oxide in a cylindrical channel. Tech Phys Lett 2014;40(3):203e6. http://dx.doi.org/10.1134/ S1063785014030067. Kuzma-Kichta YuA, Lavrikov AV, Shustov MV, Chursin PS, Chistyakova AV. Boiling characteristics at nanoparticle coated surface. In: Proceedings of the ECI 8th international conference on boiling and condensation heat transfer; 2012. Lausanne, Switzerland. Kuzma-Kichta YuA, Lavrikov AV, Shustov MV, Tsurikov D, Zhukov V. Boiling heat transfer enhancement on macro-, micro- and nano-scales. J Nucl Eng Technol 2012;2(2). Li Xiangdong, Cheung Sherman Chi Pok, Tu Jiyuan. Nucleate boiling of dilute nanofluids: mechanism of exploring and modeling. Int J Therm Sci 2014;84: 323e34. http://dx.doi.org/10.1016/j.ijthermalsci.2014.05.021. Li X, Li K, Tu J, Buongiorno J. On two-fluid modeling of nucleate boiling of dilute nanofluids. Int J Heat Mass Transf 2014;69:443e50. http://dx.doi.org/ 10.1016/j.ijheatmasstransfer.2013.10.037. Kamatchi R, Venkatachalapath S. Parametric study of pool boiling heat transfer with nanofluids for the enhancement of critical heat flux: a review. Int J Therm Sci 2015;87:228e40. http://dx.doi.org/10.1016/j.ijthermalsci. 2014.09.001. Bi J, Vafai K, Christopher DM. Heat transfer characteristics and CHF prediction in nanofluid boiling. Int J Heat Mass Transf 2015;80:256e65. http://dx.doi. org/10.1016/j.ijheatmasstransfer.2014.09.019.