Convective Heat Transfer Mechanisms and Clustering in Nanofluids

Transcription

1 2011 International Conerence on Nanotechnology and Biosensors ICBEE vol.25(2011) (2011) IACSIT ress, Singapore Convective Heat Transer Mechanisms and Clustering in Nanoluids Mohammad Hadi irahmadian Neyriz Branch, Islamic Azad University, Neyriz, Iran Abstract. Nanoluids are suspensions o nanoparticles in luids that show signiicant enhancement o their properties at modest nanoparticle concentrations. Nanoluids, due to anomalously high thermal conductivity, are important in heat transer. Regarding the various models o thermal conductivity, we show that thermal dispersion model or explaining nanoluid heat transer results have better agreement with experimental results. Investigation o the eect o nanoluid particle size and temperature on heat transer results in complicated trends due to the opposing eects o thermal conductivity and thermal dispersion on heat transer in terms o particle size dependence. Numerical results are compared with experimental and numerical data in the literature and good agreement is observed especially with experimental data. eywords: Nanoluids, Dispersion Model, Heat Transer, Boundary Condition. 1. Introduction The conventional method or enhancing heat transer in a thermal system consists o increasing the heat transer surace area as well as the low velocity o the working luid [1]. The dispersion o solid nanoparticles in heat transer luids is a relatively new method. Extended suraces such as ins and microchannels (width <100 μm) have already been used to increase the heat transer surace area. Their perormance in eectively removing as much as 1000 W/cm2 has shown a great improvement in the area o cooling. Heat transer can be enhanced by employing various techniques and methodologies, such as increasing either the heat transer surace or the heat transer coeicient between the luid and the surace, that allow high heat transer rates in a small volume. Cooling is one o the most important technical challenges acing many diverse industries, including icroelectronics, transportation, solidstate lighting, and manuacturing. Fluids behavior and spatial structure in the nanoscale undamentally are dierent rom behavior and structure in micron and larger dimensions. A nanoluid is the suspension o nanoparticles in a base luid.. Studies show that the luid has been conined or low in nanochannels has non-continuum behavior and phenomena such as lux delaminate and unconventional density luctuation appear. Furthermore, the role o interaction between lux and nano-channel is prominent and channel wall dynamic can be undamentally aect luid dynamics. With the recent improvements in nanotechnology, the production o particles with sizes on the order o nanometers can be achieved with relative ease. As a consequence, the idea o suspending these nanoparticles in a base liquid or improving thermal conductivity has been proposed recently [2,3]. 2. Improving Heat Transer eiciency 84

2 Heat transer plays an important role in numerous applications. For example, in vehicles, heat generated by the prime mover needs to be removed or proper operation. Similarly, electronic equipments dissipate heat, which requires a cooling system. Heating, ventilating, and air conditioning systems also include various heat transer processes. Heat transer is the key process in thermal power stations. In addition to these, many production processes include heat transer in various orms; it might be the cooling o a machine tool, pasteurization o ood, or the temperature adjustment or triggering a chemical process. In most o these applications, heat transer is realized through some heat transer devices; such as, heat exchangers, evaporators, condensers, and heat sinks. Increasing the heat transer eiciency o these devices is desirable, because by increasing eiciency, the space occupied by the device can be minimized, which is important or applications with compactness requirements. Furthermore, in most o the heat transer systems, the working luid is circulated by a pump, and improvements in heat transer eiciency can minimize the associated power consumption. Researches show that, there is signiicant discrepancy in nanoluid thermal conductivity data in the literature. There are several mechanisms proposed to explain the thermal conductivity enhancement o nanoluids, such as Brownian motion o nanoparticles[4], clustering o nanoparticles [5] and liquid layering around nanoparticles. Due to the lack o systematic experimental data in the literature, it is diicult to analyze the relative signiicance o these mechanisms. Most o the theoretical models based on these mechanisms include some empirical constants. The eect o parameters such as temperature, particle volume raction and particle size distribution o nanoparticles studied to heat transer enhancement with nanoluids. It is possible to correctly predict experimental results to some extent by adjusting the values o these constants accordingly. On the other hand, at present, a complete theoretical model o thermal conductivity that takes all o the parameters into account is not available. Fig.1: Schematic illustration representing the clustering phenomenon. High conductivity path results in ast transport o heat along large distances. It was shown that the thermal conductivity was aected by actors such as temperature, particle size, and ph level. Several theoretical models have been proposed to explain the behavior o nanoparticles. Many o these models can be categorized as either static or dynamic models. Static models assume that the nanoparticles are stationary in the base luid, orming a composite material. In these models, the thermal properties o nanoluids 85

3 are predicted through conductionbased models such as that o Maxwell. One such model is the modiied Maxwell theory o Hamilton and Crosser which gives the enhancement o thermal conductivity as n + 2 2( ) φ = (1) + 2 ( ) φ where n, p and are the thermal conductivity o the nanoluid, nanoparticles and base luid, respectively. φ is the volume raction o particles in the mixture. Density o nanoluids can be determined by using the ollowing expression ρ n = φρ + 1 ρ ) (2) ( And we can derived speciic heat o nanoluids by c, n φ = ( ρc ) + ( 1 φ)( ρc ) ρ n (3) 3. Clustering in Nanoluids One o the main obstacles encountered in microluid experiments was the agglomeration o particles. Even though research, such as that documented in [10, 11], shows a substantial increase in the thermal conductivity o the base luid with the addition o nanoparticles, the movement towards practical applications has been hampered by the rapid settling o the nanoparticles. The settling o particles not only decreased the overall heat transer o the luid (by decreasing the eective surace area used or heat transer), but also led to the abrasion o suraces, clogging o microchannels, and a decrease in pressure which resulted in an increase in pumping power. Although nanosized particles have greatly reduced the problem o agglomerated particles, it still occurs and can hinder the thermal conductivity o the nanoluid, especially at concentrations over 5% agglomeration is more apparent when using oxide nanoparticles because they require a higher volume concentration compared to metallic nanoparticles in order to achieve the same thermal conductivity enhancement [12]. The tendency o particles to group together beore they are dispersed in the luid is due to the van der Waals orces. This is particularly seen in metallic particles since dipoles can occur easily in the molecules o these particles. The creation o dipoles prompts the attraction o other dipoles in the vicinity. The van der Waals orces stem rom the attraction o these dipoles, which can be induced even in neutral particles. This attractive orce is considered to be the main culprit behind the agglomeration o particles, especially in nanopowders. To alleviate this problem, there have been various proposals or the manuacture and dispersion o nanoparticles in luids. One proposal involves adding surace treatments to the nanoparticles. It was seen that when copper nanoparticles were coated with a 2 10 nm thick organic layer a stable suspension would be achieved in ethylene glycol [13]. There is research currently being conducted towards improving the two-step process to produce well-dispersed nanoluids. Moreover, there exist a ew one-step processes that result in nanoparticles being uniormly dispersed and stably suspended in the base luid. One such method involves condensing copper nanopowders directly rom the vapor phase into lowing ethylene glycol in a vacuum chamber [14]. Documents [15 17] also show stable, welldispersed suspensions in nanoluids containing TiO2, CuO, and Cu. In these experiments, a one-step process called submerged arc nanoparticle synthesis was used to create the nanoparticles. Various techniques have been implemented to reduce the clustering o particles once they are in the luid [18, 19]. Usually, they involve some sort o agitation within the nanoluid to separate the clusters into individual particles and keep them rom settling. These methods include the use o dispersants, changing the ph value o the base luid, and using ultrasonic 86

4 vibration to excite the particles [16]. Among these methods, the most commonly used ones are ultrasonic vibration and the use o dispersants. Both techniques are relatively eective, but when using dispersants, the amount added to the luid must be a very low percentage (usually 1% or less). This is done so as to minimize its eects on the thermal conductivity o the nanoluid. However, it should be noted that loose particle chains may be responsible or some o the high thermal conductivities o nanoluids; see rasher et al. [19]. The Argonne National Laboratory also developed the single-step and two-step processes or the dispersion o nanoparticles in a luid [1]. The single-step process consists o simultaneously making and dispersing the particles in the luid. The two-step method separates the manuacture and dispersion o particles into two steps (particles are manuactured irst and then dispersed into the base luid). The two-step process is the more commonly used method and is usually used in conjunction with ultrasonic vibration to reduce the amount o clustered particles in the luid. Analysis o the reviewed literature shows that there is still no conclusive theory concerning the prevention o clustering in the nanoparticle suspensions. Beore using nanoluids in practical applications, the problem o clustering must be consistently kept to a minimum. When looking at longterm eects, clustering o the particles will eventually cause a decrease in the thermal conductivity o the nanoluid and may also cause wear in the pipes or pumps through which it is lowing. Thereore, nanoluids cannot be used in systems designed or long-term use until this problem is solved. Otherwise, the use o nanoluids may decrease the lie expectancy o a system, even i it improves the overall eiciency. In the mean time, an optimization and design problem persists when nanoluids are used in the ield. 4. Conclusion Convection heat transer is recognized within nanoluids, but there is not enough research results published to develop a model that ully explains this behavior in nanoluids. Furthermore, several o the research papers available seem to contradict each other as some data shows an increase in convection as the particle volume raction is increased, while other data shows deterioration in convective heat transer as the particle density and concentration were increased. Clustering still poses a problem in nanoluids even though the occurrence o agglomeration has decreased rom the previous micrometer-sized particles suspensions. Various methods are currently used to keep particles rom clustering together, but in the long run, it is inevitable. Clustering is a problem thatmust be solved beore nanoluids can be considered or long-term practical uses. Although the increase in thermal conductivity would increase the eiciency o the systems where nanoluids are used, the lie o the system may be decreased over time i particles begin to orm clusters. In igures 2 and 3 results o the numerical analysis or the hydrodynamically ully developed, thermally developing laminar low o Al O /water nanoluid inside a straight circular tube under constant wall temperature 2 3 and constant wall heat lux boundary conditions are presented. Numerical results are compared with experimental and numerical data available in the literature. Eects o particle size, heating and cooling on heat transer enhancement are investigated. It is seen that application o the thermal dispersion model to the governing energy equation provides meaningul results which are in agreement with the available experimental data in the literature [4]. This can be considered as an indication o the validity o the thermal dispersion model or nanoluid heat transer analysis. Furthermore, it can be concluded that single phase analysis o nanoluid heat transer is suiciently accurate or practical applications as long as variation o thermal conductivity with temperature is taken into account in the associated calculations [6]. Figure 2 show the parameter o particle size in the nanoluid thermal conductivity that is observed with increasing particle size, thermal conductivity increases, which results rom compliance with the experimental results is well. Investigation o the eect o nanoluid particle size on heat transer results in complicated trends due to the opposing eects o thermal conductivity and thermal dispersion on heat transer in terms o particle size dependence. It is seen that i empirical constant C in dispersed thermal conductivity expression is 87

5 suiciently small, the eect o thermal conductivity dominates particle size dependence which results in increasing heat transer with decreasing particle size. On the other hand, i C is large, heat transer increases with increasing particle size. Fig. 2: Comparison o the experimental results o the thermal conductivity ratio or Al 2 O 3 /water nanoluid. Figure 3 shows that the nanoluid heat transer revealed that or the constant wall heat lux boundary condition. Investigation o the eects o cooling and heating on nanoluid heat transer revealed that or the constant wall heat lux boundary condition, heat transer and associated enhancement is higher when low temperature is higher. When it comes to the constant wall temperature boundary condition, the dominant parameter that aects the heat transer is the wall temperature. As the wall temperature increases, the heat transer and the associated enhancement increases. These acts should be taken into account or the practical application o nanoluids in heat transer devices. Fig.3: Comparison o the experimental results o the thermal conductivity ratio or Al 2 O 3 /water nanoluid. Nanoluids have the potential to open the doors to major advancements in many high-tech industries where limits on cooling have posed limits on innovation. Since all other cooling options have been exhausted, nanoluids are the only option let with the possibility o increasing heat transer capabilities o current systems. However, a ull understanding o the mechanisms behind the enhancement o thermal conductivity in nanoluids has not been reached and there is still disagreement between some o the experimental results. This lack o agreement has led to the generation o various models. Once a general model that ully explains the behavior o nanoparticle suspensions has been developed, steps can be taken towards practical uses. Moreover, better techniques or the dispersion o particles in luids must be created so as to minimize clustering. When these objectives have been reached, nanoluids will enter the practical arenas o science in a more meaningul way. 5. Reerences 88

6 [1] S.. Das, S. U. S. Choi, W. Yu, and T. radeep, Nanoluids:Science and Technology, John Wiley & Sons,NJ, [2] H. Masuda, A. Ebata,. Teramae, and N. Hishinuma, Netsu Bussei, 4(4), (1993), [3] S. U. S. Choi, The American Society o Mechanical Engineers, New York, FED-Vol. 231 / MD-Vol.66, (1995), [4] Y. Li, W. Qu, and J. Feng, Chinese hys. Lett., 25(9), (2008), [5] S.. Jang, and S. U. S.Choi, Appl. hys. Lett., 84(21), (2004), [6] R. rasher,. Bhattacharya, and. E. helan, hys. Rev. Lett., 94(2), (2005), [7] C. H. Chon,. D. ihm, S.. Lee, and S. U. S. Choi, Appl. hys. Lett., 87(15), (2005), [8] S.. Das, N. utra,. Thiesen, and W. Roetzel, J. Heat Transer, 125(4), (2003), [9] C. H. Li, and G.. eterson, J. Appl. hys., 99(8), (2006), [10] S. Lee, S. U.-S. Choi, S. Li, and J. A. Eastman, Measuring thermal conductivity o luids containing oxide nanoparticles, Journal o Heat Transer, vol. 121, no. 2, pp , [11] H. Xie, J.Wang, T. Xi, Y. Liu, F. Ai, and Q.Wu, Thermal conductivity enhancement o suspensions containing nanosized alumina particles, Journal o Applied hysics, vol. 91, no. 7, pp , [12] W. Yu,D.M. France, J. L. Routbort, and S. U. S. Choi, Review and comparison o nanoluid thermal conductivity and heat transer enhancements, Heat Transer Engineering, vol. 29, no. 5, pp , [13]. S. Suslick, M. Fang, and T. Hyeon, Sonochemical synthesis o iron colloids, Journal o the American Chemical Society, vol. 118, no. 47, pp , [14] J. A. Eastman, S. U. S. Choi, S. Li, W. Yu, and L. J. Thompson, Anomalously increased eective thermal conductivities o ethylene glycol-based nanoluids containing copper nanoparticles, Applied hysics Letters, vol. 78, no. 6, pp , [15] H. Chang, T. T. Tsung, Y. C. Yang, et al., Nanoparticle suspension preparation using the arc spray nanoparticle synthesis system combined with ultrasonic vibration and rotating electrode, International Journal o Advanced Manuacturing Technology, vol. 26, no. 5-6, pp , [16] C.-H. Lo, T.-T. Tsung, L.-C. Chen, C.-H. Su, and H.-M. Lin, Fabrication o copper oxide nanoluid using submerged arc nanoparticle synthesis system (SANSS), Journal o Nanoparticle Research, vol. 7, no. 2-3, pp , [17] C.-H. Lo, T.-T. Tsung, and L.-C. Chen, Shape-controlled synthesis o Cu-based nanoluid using submerged arc nanoparticle synthesis system (SANSS), Journal o Crystal Growth, vol. 277, no. 1 4, pp , [18] Q. Cao and J. Tavares, Dual-lasma Synthesis o Coated Nanoparticles and Nanoluids, November 2006, [19] S. M. S. Murshed,. C. Leong, and C. Yang, Thermophysical and electrokinetic properties o nanoluids a critical review, Applied Thermal Engineering, vol. 28, no , pp ,