xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002069377
007 cr mnu|||uuuuu
008 100421s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003237
Martinez, Christian David.
Heat transfer enhancement of spray cooling with nanofluids
h [electronic resource] /
by Christian David Martinez.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 57 pages.
Thesis (M.S.M.E.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: Spray cooling is a technique for achieving large heat fluxes at low surface temperatures by impinging a liquid in droplet form on a heated surface. Heat is removed by droplets spreading across the surface, thus removing heat by evaporation and by an increase in the convective heat transfer coefficient. The addition of nano-sized particles, like aluminum or copper, to water to create a nanofluid could further enhance the spray cooling process. Nanofluids have been shown to have better thermophysical properties when compared to water, like enhanced thermal conductivity. Although droplet size, velocity, impact angle and the roughness of the heated surface are all factors that determine the amount of heat that can be removed, the dominant driving mechanism for heat dissipation by spray cooling is difficult to determine. In the current study, experiments were conducted to compare the enhancement to heat transfer caused by using alumina nanofluids during spray cooling instead of de-ionized water for the same nozzle pressure and distance from the heated surface. The fluids were sprayed on a heated copper surface at a constant distance of 21 mm. Three mass concentrations, 0.1%, 0.5%, and 1.0%, of alumina nanofluids were compared against water at three pressures, 40psi, 45psi, and 50psi. To ensure the suspension of the aluminum oxide nanoparticles during the experiment, the pH level of the nanofluid was altered. The nanofluids showed an enhancement during the single-phase heat transfer and an increase in the critical heat flux (CHF). The spray cooling heat transfer curve shifted to the right for all concentrations investigated, indicating a delay in two-phase heat transfer. The surface roughness of the copper surface was measured before and after spray cooling as a possible cause for the delay.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Frank Pyrtle III, Ph.D.
Critical heat flux
x Mechanical Engineering
t USF Electronic Theses and Dissertations.
Heat Transfer Enhancement of Spray Cooling with Nanofluids by Christian David Martinez A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Mechanical Engineering Department of Mechanical Engineering College of Engineering University of South Florida Major Professor: Frank Pyrtle III, Ph.D. Muhammad M. Rahman, Ph.D. Craig Lusk, Ph.D. Date of Approval: November 3, 2009 Keywords: nanoparticles, thermal management, critical heat flux, al umina, phase change Copyright 2009, Christian David Martinez
Dedication I would like to dedicate th is thesis to my parents b ecause without them I would not have been able to accomplish this. I want to dedicate this to my dad because everyday he teaches me what is possible wh en you work hard and never look back, ni para cojer empurso. Also to my mom, becau se she always believed in me and knew that I could do it, even if sometimes I doubted it myself. Thank you for telling me to keep going, over and over again. Im doing good, good now.
Acknowledgements I would like to thank Dr. Pyrtle for his a ssistance in completing this thesis. Thank you for allowing me work with you under the Research Experience for Undergraduates program. Your heat transfer class was one of the main reasons I wanted to pursue a graduate degree. Thank you for all your help and knowledge. I would also like to thank Dr. Lusk and Dr. Rahman for being part of my committee. I would like to thank my lab partners John Shelton and Elliott Rice. Thank you John for helping me understand heat transfer better througho ut the years and helping me assist students with their LabVIEW labs. Also, thank you for your assistance in the writing of this thesis. I woul d like to thank Elliott for having the same drive as I do to always be the best. Thank you for st udying with me for long hours and for your assistance during my thesis work. I would like to thank Karen Mann for listeni ng to me when I tried to explain what was going on with the spray cooling experiment s and for helping me converting the thesis from Word documents into a single .pdf file. Thank you all.
i Table of Contents List of Tables iii List of Figures iv Abstract vi Chapter 1 Introduction 1 Chapter 2 Objective of Current Study 4 Chapter 3 Literature Review 5 3.1 Nanofluids 5 3.1.1 Effects of pH on Nanofluids 6 3.2 Heat Transfer with Nanofluids 7 3.2.1 Transient Hot Wire Method Research 8 3.2.2 Pool Boiling Research 11 3.2.3 Impinging Jet Research 16 3.2.4 Spray Cooling Research 18 Chapter 4 Experimental Set-up and Procedure 22 4.1 Nanofluid Preparation 22
ii 4.2 Copper Block 24 4.3 Spray System 27 4.4 Spray Surface Preparation 28 4.5 Acquisition System 29 4.6 Surface Roughness Measurement 30 4.7 Experimental Procedure 31 Chapter Five Results and Discussion 32 5.1 Uncertainty Analysis 32 5.2 Experimental Results 34 Chapter Six Conclusion and Recommendations 53 6.1 Conclusion 53 6.2 Recommendations 54 References 55
iii List of Tables Table 1: Summary Table 19 Table 2: Properties of Aluminum Oxide Nanoparticles 22 Table 3: pH Level of Se lected M ass Concentrations 23 Table 4: Mass Flow Rates 28 Table 5: Critical Heat Flux for Water 37 Table 6: Critical Heat Flux for 1.0% wt. Alumina Nanofluids 40 Table 7: Critical Heat Flux for 0.5% wt. Alumina Nanofluid 43 Table 8: Critical Heat Flux for 0.1% wt. Alumina Nanofluid 45
iv List of Figures Figure 1: pH Level vs. Mass Concentration of Alumina Nanofluids 23 Figure 2: Copper Block Design 25 Figure 3: Boundary Temperature Profile 26 Figure 4: Heat Flux Path through Block 26 Figure 5: Heat Flux Normal to Spray Surface 27 Figure 6: Schematic of Spray System 28 Figure 7: LabVIEW Front Panel 30 Figure 8: Spray Cooling Curve for W ater at 40 Psi 35 Figure 9: Spray Cooling Curve for W ater at 45 Psi 36 Figure 10: Spray Cooling Curve for Water at 50 Psi 36 Figure 11: Spray Cooling Curve Comparison of Water at Different Pressures 37 Figure 12: Spray Cooling Curve for 1.0% wt. AluminaNanofluid at 40 Psi 38 Figure 13: Spray Cooling Curve for 1.0% wt. Alumina Nanofluid at 45 Psi 39 Figure 14: Spray Cooling Curve for 1.0% wt. Alumina Nanofluid at 50 Psi 39 Figure 15: Spray Cooling Curve Comparis on for 1.0% wt. Alum ina Nanofluids at Different Pressures 40 Figure 16: Cooling Curve for 0.5% wt. Alumina Nanofluid at 40 Psi 41 Figure 17: Spray Cooling Curve for 0.5% wt. Alumina Nanofluid at 45 Psi 41 Figure 18: Spray Cooling Curve for 0.5% wt. Alumina Nanofluid at 50 Psi 42
v Figure 19: Spray Cooling Curve Comparis on for 0.5% wt. Alum ina Nanofluids at Different Pressures 42 Figure 20: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 40 Psi 43 Figure 21: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 45 Psi 44 Figure 22: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 50 Psi 44 Figure 23: Spray Cooling Curve Comparis on for 0.1% wt. Alum ina Nanofluid at Different Pressures 45 Figure 24: Spray Cooling Curve Comparison of Water vs. Nanofluids at 40 Psi 46 Figure 25: Spray Cooling Curve Comparison of Water vs. Nanofluids at 45 Psi 46 Figure 26: Spray Cooling Curve Comparison of Water vs. Nanofluids at 50 Psi 47 Figure 27: Surface Roughness before Spray Cooling 49 Figure 28: Surface Roughness after Spray C oolin g with 0.5% wt. Alumina Nanofluid 50 Figure 29: Surface Roughness after Cleaning Procedure 50
vi Heat Transfer Enhancement of Spray Cooling with Nanofluids Christian David Martinez ABSTRACT Spray cooling is a technique for achie ving large heat fluxes at low surface tem peratures by impinging a liquid in droplet form on a heated surface. Heat is removed by droplets spreading across the surface, thus removing heat by evaporation and by an increase in the convective heat transfer coeffi cient. The addition of nano-sized particles, like aluminum or copper, to wa ter to create a nanofluid could further enhance the spray cooling process. Nanofluids have been show n to have better ther mophysical properties when compared to water, like enhanced thermal conductivity. Although droplet size, velocity, impact angle and the roughness of the heated surface are all factors that determine the amount of heat that can be removed, the dom inant driving mechanism for heat dissipation by spray cooling is difficult to determine. In the current study, experiments were conducted to compare the enhancement to heat transfer caused by using alumina nanof luids during spray cooling instead of deionized water for the same nozzle pressure and distance from the heated surface. The fluids were sprayed on a heated copper surf ace at a constant distance of 21 mm. Three mass concentrations, 0.1%, 0.5%, and 1.0%, of alumina nanofluids were compared against water at three pressure s, 40psi, 45psi, and 50psi. To ensure the suspension of the
vii aluminum oxide nanoparticles during the experiment, the pH level of the nanofluid was altered. The nanofluids showed an enhancem ent during the single-phase heat transfer and an increase in the critical heat flux (C HF). The spray cooling heat transfer curve shifted to the right for all concentrations investigated, indicating a delay in two-phase heat transfer. The surface roughness of the copper surface wa s measured before and after spray cooling as a possible cause for the delay.
1 Chapter 1 Introduction Gases or liq uids impinging on a flat surface have been used to enhance the heating, cooling, or drying of a surface due in part to the increase in convection heat transfer coefficient. The delivery of the ga s or liquid to the surface has been achieved by the use of a single nozzle or an array of nozzles usually oriented normal to the target surface. Impinging jets have been used in many applications including the annealing of metals and the cooling of gas turbine blades. One particularly impor tant application of impinging jets is the spray cooling of high performance electronic devices (Incropera, DeWitt, Bergman, and Lavine 402). The need for these electronic devices to be smaller and faster requires the remova l of large heat fluxes to keep the product working and extend its life cycle. Currently, many electronic devices use a heat sink and fan combination to remove heat because of th eir simplicity and low cost. The heat sink conducts heat from the heated surface effici ently because of its high thermal conductivity and dissipates the heat through its fins to the surroundings via forced convection using a fan when is usually mounted on top of the heat sink. Another popular way to remove heat is by the use of heat pipes. Heat pi pes most commonly use the evaporation of water or some kind of coolant to remove heat from a heated surface. The hot end of the heat pipe vaporizes the working fluid increasing th e vapor pressure at that end providing the driving force needed to move the vapor to the cooler condensi ng end and providing the hot end with the lower temperatur e working fluid once again.
2 Both the heat sink and the heat pipe, though th ey are widely used, ha ve their limitations. To be effective at removing and the spread ing the heat, the heat sink needs to be a number of times larger than the heated su rface, usually a computer processing unit or CPU, leading to a size constraint of the electr onic device. Also, parts of the heat sink, like the pins, that are father away from the heat source are, by nature, cooler which reduces the rate of heat transf er. Heat pipes suffer from different limitations. Since most heat pipes depend on pressure differences to remove heat, the in teraction between the liquid and vapor phases can cause the heat transfer rate to deteriorate because of pressure losses caused by entrainment. To remove larg e amounts of heat with heat pipes requires longer distances to avoid vaporizing all the liqui d in the heat pipe rendering it useless. One way to remove large amounts of heat from CPUs and other similarly heated surfaces without the need for long distances or large pieces of metals and fans is with spray cooling. Spray cooling typically invol ves the phase change heat transfer of a liquid to a vapor by impingement on a flat heated surface. The most common fluid used is water because of its well known thermal properties, ab undance, cost effectiveness, easiness to store and its harmlessness to the environment. Typically, the water is delivered to the surface in a mist through the use of a round or rectangular nozzle. The enhancement for removing large quantities of h eat comes from the increased value of the convection heat transfer coefficient. The convection heat-tra nsfer coefficient during spray cooling varies not only with the temperature between the surf ace and the fluid but also with the sprays characteristics. The sprays characteristics include but are not limited to: temperature and thermal conductivity of the water, droplet si ze, velocity and angle. If the thermal
3 properties of the water were to be enhanced then, theoretically, that should lead to an enhancement of convection heat transfer coe fficient and increase the heat that can be removed from the surface. One way to change the thermal physical properties of water is by the addition of nano-size particles to create a nanofluid [C hoi]. Research on nanofluids has shown an increase in the thermal conductivity over the ba se fluid alone [Choi]. The increase in the thermal conductivity of water ha s the potential to enhance th e heat flux removed from a heated surface during spray cooling by increasing the convective heat transfer coefficient. There are other properties that can affect the effectiveness of spray cooling using nanofluids, like the surface roughness of th e heated surface, that also need to be investigated.
4 Chapter 2 Objectives of Current Study The objective of the curren t study is to determine the effectiveness of alumina nanofluids for dissipating heat from a heated copper surface using a lateral spray cooling experiment. The data collected is compared to de-ionized water at the same nozzle pressure and distance from the surface. Different mass concentrations of alumina nanofluids at different pressures will be comp ared to attempt to establish an optimum combination of concentration and pressure. Other parameters can have an effect on the effectiveness of spray cooling, such as th e surface roughness of the impinged surface. Therefore, the surface roughness of the copper su rface is recorded before and after spray cooling with the alumina nanofluid to invest igate the effects of the nanoparticles on the copper surface.
5 Chapter 3 Literature Review 3.1 Nanofluids There are many different types of nanoflu ids that can be made by using different nanoparticles and base fluid combinations. Some of the most common nanoparticles used are Alumina Oxide (Al2O3), Copper II Oxide (CuO), Zinc Oxide (ZrO2), and Silica Oxide (SiO2). The most common base fluids used for nanofluids are de-ionized water and ethanol. All nanofluids follow a basic preparation technique. Once the desired weight or volume fraction has been determined, the nanopa rticles are added into the base fluid and mixed. Mixing is usually done by ultrasonicatio n to avoid settling of the particles. The amount of time spent mixing the nanofluids de pends on the many factors such as the ratio of base to nanoparticles, how long the experiment will last, and the weight or volume fraction used. The results of the first research into nanofluids conducted by Choi et al. (1995) showed that these new nanofluids had tremendous heat transfer applications because of their improved heat transfer properties. A lo t of research has gone into finding exactly why nanoparticles have such enhancement to h eat transfer properties of the fluid but no definitive answers have been found. Jang et al. (2004) and Chon et al. (2005) have theorized that microconvection induced by Br ownian motion of the nanoparticles is one
6 of the driving mechanisms behind the thermal enhancements of nanofluids. The random motion of the nanoparticle s would create a source of fluid convection that would increase the thermal properties of the base fluid. Most researcher s agree that nanofluids have been shown experimentally to have better heat transfer propertie s than the base fluid alone. Another advantage of utilizing nanofluid s is that at the nano-scale the particles are small enough to stay in suspension, under the ri ght conditions they can stay in suspension indefinitely, effectively eliminati ng sedimentation, clumping, and clogging. 3.1.1 Effects of pH on Nanofluids One of the most common challenges in using nanofluids is maintaining the suspension of the nanoparticle s within the fluid. Anoop et al. (2009) was able to accomplish suspension of aluminum oxide partic les for several weeks by altering the pH value of the nanofluid. By keeping the nanoflu id away from the iso-electric point (IEP), the point where there is zero net charge between the particles and the bulk fluid, the particles were kept in suspen sion by the electrostatic repulsive forces between them. The pH values of 1 wt%, 2 wt%, 4 wt% and 6 wt% were found to be 6.5, 6, 5.5, and 5 respectively. The dispersion behavior a nd thermal conductivity of Al2O3 water nanofluids under different pH levels were investigated by Zhu et al. (2009). For all the experiments a 0.1 wt% alumina nanofluid concentration was used. To control the pH level of the nanofluid Zhu et al. used analytical grade hydrochloric acid (HCl) and sodium hydroxide (NaOH). To aid in the initial dispersion of the nanoparticles an ioni c surfactant, sodium dodecylbenzenesulfonate (SDBS), was added to the mixture and then mixed in an
7 ultrasonicator. Zhu et al. found that for an alumina na nofluid containing SDBS as a surfactant, the optimum pH value is 8.0. This is the point with the greatest value of zeta potential and therefore the particles have the highest electrostatic repulsive forces, which keep the particles in suspen sion. The thermal conductivity of the alumina nanofluid was measured by the transient plane source (T PS) method. Through the investigation it was found that there is an increas e in thermal conductivity for pH values from 3.0 to 8.0-9.0. Zhu et al. suggest that as the pH le vel of the nanofluid increas es farther away from the point of zero change (PZC), the point wher e there are no repulsive forces between the Al2O3 nanoparticles, therefore th ey coagulate. As a resu lt, the hydration forces are greater between the particles. The increase in hydration forc es causes an enhancement in the mobility of the nanoparticles. The mob ility of the nanoparticle s creates microscopic motions that cause microconvection which enhances the heat transfer process. 3.2 Heat Transfer Research with Nanofluids Its been shown that nanofluids in genera l have better heat transfer properties than the base fluid alone, specifically better therma l conductivity and heat transfer coefficient. These heat transfer properties theoretica lly should make nanofluids ideal for phase change heat transfer processes. These enhancements have been researched using experiments such as the transient hot wire method, pool boiling, impinging jet and nanofluid tube flow.
3.2.1 Transient Hot Wire Method Research The transient hot wire method (THW) is a transient dynamic technique where the temperature rise of a sample is measured at a defined distance from a heat source. The hot wire is assumed to have a uniform heat output along its length and the thermal conductivity of the sample can be calculated from the temper ature change of the sample over a known time interval. The thermal conductivity of different concentrations of water-copper and transport oil-copper nanofluids were investigated by Xuan et al. (2000) by the use of the transient hot wire method. To calculate the thermal conductivity of the nanofluids, Xuan et al. used the fundamental e quation of the transient hot wire method, give by: C r at k q trT24 ln 4 , where k is the thermal conduc tivity of the sample, a is the thermal diffusivity, and C is given by: geC where g (g = 0.5772157) is Eulers constant. The re sults show that one of the factors affecting the thermal conductivity of nanofluids is the nanoparticle volume fraction. An increase in volume fraction results in an increase in the thermal conductivity of both the water-copper and the transformer oil-copper na nofluids. For example, the water-copper nanofluid saw an improvement in the therma l conductivity ratio of nanofluid to water from 1.24 to 1.78 with an increase of volume fraction of 2.5% to 7.5%. Hwang et al. (2006) also investigated th e effects of nanoparticle c oncentration on the thermal conductivity of nanofluids us ing the THW method. The investigation was conducted 8
9 with multiwalled carbon nanotubes (MWCNT) in water, copper monoxide (CuO) in water, silicon dioxide (SiO2) in water, and CuO in ethynele glycol. The results of the investigation were similar to Xuan et al., where an increase in the thermal conductivity of the nanofluids was obtained with an increase in the volume fraction concentration of the nanoparticles. Hwang et al. also reported that the ther mal conductivity of nanofluids were also dependent on the thermal conductivity of the nanoparticles and the base fluid. For instance, for the same volume fraction co ncentration of 1% th e CuO-water nanofluid saw an increase in the thermal conductivity of approximately 5% when compared to an improvement of approximately only 3% for SiO2-water nanofluids. One possible factor for the difference in improvement is the th ermal conductivity of th e nanoparticles, 76.5 W/mK for CuO compared to only 1.38 W/mK for SiO2. Different enhancements in thermal conductivity where also acquired for nanofluids with the same nanoparticles but different base fluids. The enhancement to thermal conductivity for CuO-ethynele glycol nanofluids was higher than that for CuO-wa ter nanofluids for the same volume fraction concentration. The results show that the ba se fluid with the lowest thermal conductivity will benefit more from the addition of nanopartic les, in this case the ethynele glycol with a thermal conductivity of 0.252 W/mK compared to that of water with 0.613 W/mK. Zhang et al. (2006) used a method based on the TH W method called the short hot wire (SHW) method to conduct experiments with different nanoparti cle and base fluid combinations. Different concentrations of nanoparticles and the temperature of the nanofluid are investigated for their effects on the thermal conductivity of the nanofluid. In the study gold (Au)-toluene nanofluid at a volume fraction of 0.003%, Al2O3-water nanofluids with mass concentrations of 0%, 10%, 20% and 40%, and carbon nanofiber
(CNF)-water nanofluids with a volu me concentra tion range of 0 to 1% are investigated. Zhang et al. also recorded increases in thermal conductivity of all nanof luids investigated corresponding to increases in the concentration of the nanopa rticles and the temperature of the nanofluid. The slope of the depende nce of the thermal conductivity on temperature for nanofluids was compared to pure water and it was found that the slopes were the same. The results indicate that the temper ature dependence on the thermal conductivity and thermal diffusivity of the nanoparticle s do not have an affect on the thermal conductivity and thermal diffusivity of the na nofluid for the given concentrations. Xie et al. (2002) also used the THW to study the th ermal conductivity of nanofluids by looking at different volum e fractions of Al2O3 particles suspended in de -ionized water, ethanol, and pump oil, different specific surface areas, and by looking at the different pH values of the nanofluid. Xie et al. found that for all the base fluids the thermal conductivity increases with increasing volume fraction but with different slopes, corresponding to different pH values. The results show that with an increase in pH level the enhanced thermal conductivity ratio decreases. When th e difference between the pH value of the suspension and the isoelectric point increas es, the hydration forces among the particles start to increase which leads to an enhancem ent of the mobility of the nanoparticles in the fluid. This enhancement in the mobility of nanoparticles causes microconvection that enhances the heat transfer process. The re sults show that there is an optimum specific surface area of the nanopartic les that enhance thermal conductivity. The thermal conductivity increases with increasing specific surface area at first but then begins to decrease. The optimum specific surface area for this study is found to be 25 One of the factors for this change in therm al conductivity is that as the particle size of the 12 gm 10
11 nanoparticle decreases, the specific surface ar ea increases proportionally. Since heat transfer in nanofluids occur at the particle-fluid interface, a reduction in particle size can result in a large inte rfacial area. Murshed et al. (2005) prepared nanofluids by dispersing titanium oxide (TiO2) nanoparticles in rod and spherical shapes in de-ionized water to conduct THW experiments. The results show that the thermal conductivity increases with increasing nanoparticle volume concentrat ion. The shape of th e nanoparticles also affects the thermal conductivity of th e nanofluid. The rod shaped TiO2 nanoparticles showed an enhancement of 33% in thermal c onductivity when compared to the base fluid alone at a volume concentration of 5%. In comparison, the spherical shaped nanoparticles showed an enhancement of 30% at the same volume concentration. 3.2.2 Pool Boiling Research Pool boiling is the process in which vapor is created at the liquid-surface interface by a surface heated above the satu ration temperature of the bulk fluid. The motion of the vapor and the surrounding fluid near the heat ed surface is due to buoyancy forces. As vapor escapes the surface, liquid comes in to fill the void and this process removes heat from the heated surface. Bang et al. (2005) investigated the boiling heat tr ansfer characteristics in different volume concentrations of alumina nanofluids and compared the results to pure water. Both vertical and horizontal heated surfaces were considered for the experiment. The research shows that the additi on of alumina nanopart icles causes the boi ling curve to shift to the right, which means that there are de creases in the pool nucleate boiling heat transfer for all concentrations. Also, it wa s observed that the nucleate boiling regime was
12 delayed due to an extended natural convec tion stage which is inconsistent with the increase in thermal conductivity of nanofluids. On the other hand, the critical heat flux (CHF) was increased by 32% and 13% for horizon tal and vertical heaters respectively. Bang et al. suggested that the fouling of the h eated surface by the alumina nanoparticles caused a decrease in the nucleation site density Large vapor blankets close to the surface are generated with the decrease in nuclea tion sites which allows more water to be supplied to the heated surface. You et al. (2003) conducted pool boiling experiments of Al2O3 water nanofluids at a pressu re of 2.89 psia which gives a saturation temperature of 60 C using a 1 x 1 cm2 polished copper surface. The na noparticle mass concentrations ranged from 0 g/l to 0.05 g/l and their results were compared to de-ionized water. The results show an increase in the CHF with an increase of mass concentration. A remarkable increase of 200% enhancem ent was shown with a 0.05 g/l mass concentration. Another result of the study shows that the average size of the bubbles increased and the frequency decreased with the use of nanofluids. You et al. concludes that the increase in the CHF is not related to the increase in thermal conductivity by the addition of nanoparticles. Das et al. (2003) also investigated the boiling heat transfer characteristics of 1%, 2%, and 4% concentra tion alumina nanofluids with similar results to those obtained by Bang et al. The boiling curve again showed a shift to the right with increasing concentration of nanoparticles. Das et al. considered the surface roughness of the heaters as one of the factors for the de grading of the heat transfer performance. Surface roughness measurements of the heated copper surface showed that after pool boiling experiments with nanoflu ids, the surface of the heater was smoother than before the experiment. The results suggest that th e alumina nanoparticles are being trapped on
13 the surface, since the size of the particles we re one to two orders of magnitude smaller than the roughness. The trapped particles form a layer on the heated surface hindering fluid flow and heat transfer, which may expl ain the degrading of boiling heat transfer performance when compared to water. Das et al. again investigated 1%, 2%, and 4% concentration of alumina nanofluids on pool boiling but on narrow horizontal tubes. The tubes were 4 and 6.5 mm in diameter. Once again, a deterioration of the pool boiling heat transfer curve resulted with increasing nanoparticle concentration. The deterioration was less significant for the narrow tubes than t ubes of a larger diameter (20 mm). Less deterioration in the narrow tubes was believed to be due to the change in bubble diameter and sliding bubble mechanism when compared to the larger diameter tubes. Das et al. concluded that there are tw o conflicting phenomena occurring with pool boiling heat transfer with nanofluids. The addition of nanoparticles increases the viscosity of the base fluid which increases the heat transfer of the base fluid but it is overshadowed by the decrease in the nucleation site density due to nanoparticles impinging on the surface. Zhou et al. (2004) conducted pool boiling experiment s with different concentrations of Cu-acetone nanofluids and with acoustic cav itations. Cavitations are the sudden formation and collapse of low-pressure bubbles due to mechanical forces. In this experiment ultrasound was created by an ultrasonic vibrator. Acoustic cavitations enhance heat transfer by util izing the energy released by the collapsing low-pressure bubbles. An increase in single-phase h eat transfer was found with increasing concentration. Though a degrading of pool boiling heat transfer was found, Zhou et al. noted that when the concentration increas ed from 0.133 g/l to 0.267 g/l no further degrading was noticed. This result is substa ntially different than the work done by Das et
14al. who found that heat transfer reduced with increasing nanoparticle concentration. The thermophysical properties of the nanopart icles are though to be a reason for this discrepancy. At all the nanopa rticle concentrations investigated the acoustic cavitations were shown to enhance heat transfer. As the distance between the sound source and the heated copper surface increased from 20 mm to 40 mm, only a slight decrease in pool boiling heat transfer was noticed. Differen t volume fractions of a different nanofluid, titanium dioxide and the refrigerant HCFC 141b, was investigated by pool boiling by Trisaksri et al. (2009). The investigation used 0.01, 0.03, and 0.05 vol% of TiO2 and a cylindrical copper tube as the boiling surface. The first results from the experiment reveal that for the 0.01 vol% concentration the boiling heat transfer is the same as the base fluid alone. This shows that adding ve ry small amounts of nanoparticles to the base fluid had no effect on boiling heat transfer. At 0.03 and 0.05 vol% concentration the boiling curve is shifted to th e right indicating a deteriora tion of boiling heat transfer, which supports the results by Bang et al. One explanation for the shift of the boiling curve is the range of the excess temperatur e in the natural convection regime of the nanofluid is larger than that for the base fluid alone; this causes a delay of nucleate boiling and a rise in the surf ace temperature. Trisaksri et al. also looked at the effects of pressure on the heat transfer coefficient. At lower concentrations, 0.01 and 0.03 vol%, the effects of pressure on heat transfer coefficient are negligible. However, at 0.05 vol% there is a rise in the heat transf er coefficient at high heat fluxe s. The rise in heat transfer coefficient is lower than the rise seen for the base fluid alone. Wen et al. (2008) conducted a pool boiling experiment using different particle concentrations in alumina nanofluids with different results. The results show that there is an enhancement of both
15 boiling heat transfer coeffici ent and thermal conductivity when compared to the base fluid. The improvement increases with increa sing nanoparticle concen tration and is more significant at higher heat fluxes. Enhancement of up to 40% in heat transfer coefficient was achieved with a concentration of 1.25 wt%. With an increase of 10% with a concentration of 1.6%, the enhancement to the thermal conductivity was not as significant as for the heat transfer coefficient. Wen et al. suggests nanoparticle migration as one of the reasons for the enhancement in heat transfer coefficient and thermal conductivity and the depositing of nanoparticle s on the heated surface, which introduces a thermal resistance, as one of the reasons for the deterioration that has been seen in other studies. Vassallo et al. (2004) pool boiling experiment was done using silica oxide nanofluids with different particle sizes. In this experiment there was no decrease in the heat transfer coefficient, but no improveme nt was found either. The boiling curve for both particle sizes, 15nm and 50nm, follow the pure water boiling curve through the nucleate boiling regime. Again, an increase in the CHF was found. Coursey et al. (2008) researched an improvement in surface wetta bility as the possible mechanism for the increase in CHF. Wetting is the ability of a fluid to remain in contact with a solid surface. It was found that na noparticles had a positive effect when there was a large contact angle between the fluid and the solid surface, which means that the surface is difficult to wet or the base fluid is less wett ing. For fluids that are naturally more wettable, for example ethanol, the addition of nanoparticles had lit tle to no effect on wetting. Water had increase in wetting with the addition of nanopart icles because it is a naturally less wettable fluid. The increase in wetting was found to be one of the driving mechanisms to improving the CHF. The conflic ting results in heat transfer and thermal
16 conductivity by the addition of nanoparticles to a base flui d shows that the further research is needed in this field. 3.2.3 Impinging Jet Research Impinging jet research is another way to study the effects that nanoparticles have on the heat transfer coefficients of the base fl uids. A nozzle is used to spray a jet of fluid onto a heated surface to enhance the heat tran sfer coefficients for convective heating, cooling, or drying. Nguyen et al. (2009) used a nozzle with a diameter of 3mm to spray a 36nm alumina nanofluid onto a confined and submerged heated aluminum surface. Nguyen et al. tested different concentrations of nanopa rticles, 0%, 2.8%, and 6%, with different flow rates and nozzle-to-surface distances. Th e research shows that in some cases the addition of nanoparticles increases the heat transfer coefficient of the base fluid. With a mass flow rate of 0.15 kg/s and a nozzle-to-s urface distance of 2mm, the pure water has the highest heat transfer coefficient follo wed by 2.8% concentration and finally 6% concentration. With the same mass flow ra te but with a distan ce of 5mm, the 2.8% concentration of nanoparticles was found to give the highest heat transfer coefficient followed by water then 6% concentration. At 10mm nozzle-to-surface distance, water and 2.8% concentration have almost the same heat transfer coefficient, while the 6% concentration came in at third place. The study shows that there is an optimum nanoparticle concentration, flow rate, and nozzle-to-surface distance that will give the best results. Also, concentr ations greater than 6% shoul d be avoided for impinging jet cooling. Liu et al. (2007) conducted impinging jet research using CuO nanoparticles
17 suspended in water. The effects of nanopart icle concentration and the flow conditions were investigated and compared to the ba se fluid. The impingement took place in a 20mm diameter heated c opper surface with a 4mm diameter nozzle and the mass concentrations of CuO nanoparticles changed fr om 0.1 to 2 wt%. The results of the study show that the jet boili ng curves for all nanoparticle concentr ations are shifted to the right, indicating a deterioration of bo iling heat transfer when compared to the base fluid. For the range tested, the different nanoparticle con centrations had little effect on the boiling heat transfer. At higher jet velocities, as e xpected, the boiling heat transfer increases. The critical heat flux (CHF) of the nanofluid s increased, up to 25% compared to water, with increasing concentrations at a low ra nge. At 1 wt% no more increase in CHF was noticed. Liu et al. conducted surface roughness meas urements before and after impinging jet with the base fluid and the na nofluid. After the water jet impingement experiments were conducted the surface had be come slightly oxidized. The existence of a thin sorption layer was present after the na nofluid impingement test The sorption layer made the copper heater surface smoother, thus decreasing the number of nucleation sites. The sorption layer could explain the decrease in boiling heat transfer and the increase in CHF. The decrease in nucleation sites and the increase in thermal resistance caused by the sorption layer could be a reason for the decrease in boiling heat transfer. The existence of the sorption layer also enhances the trapping of liquid in the porous layer and prevents vapor blankets from forming leading to an increase in CHF.
183.2.4 Spray Cooling Research Another method that utilizes the impinge ment of a working fluid onto a heated surface is spray cooling. During spray cooli ng the pressure difference between the nozzle and the environment is sufficient to creat e droplets of the working fluid and those droplets impinge the surface to remove heat. Shen (2009) investigated the hydrodynamic characteristics of droplets impinging on a polished and a nano-structured heated su rface. The results of a single-wall-carbonnanotube nanofluid were compared to water. The addition of na noparticles resulted in larger spreading velocities, larger spreadi ng diameters, and an increase in early stage dynamic contact angle. It wa s found that the evaporation time was reduced by 37% with the use of nanofluids on the polished surface. The combination of the nanofluid and the nano-structured surface yielded reduced evapor ation times of 20%. The reduction of the evaporation time indicates an enhancement to heat transfer for evaporative cooling. Coursey (2007) has added high aspect ratio microchannels to the copper sprayed surface resulting in very high enhancements. An enhancement of 200% was noticed in the single-phase regime and since the two-phase regime was delayed, a heat transfer enhancement of up to 181% was achieved. In terestingly, the onset of the two-phase regime was found to occur at a temperature th at was independent of the nozzle pressure and mass flow rate. Duursma et al. (2009) conducted an inve stigation of the droplet impinging mechanics using dimethyl sulfoxide (DMSO) and ethanol nanofluids. The nanoparticles used in the inve stigation where aluminum with mass concentrations of up to 0.1% for DMSO and 3.2% for ethanol Single droplets where impinged onto the surface where high-speed photographic images we re taken to show the differences in
19 droplet behavior. The results revealed that droplet mechanics are mostly a function of Weber number and excess temperature. An in crease in the nanopart icle concentration results in a decrease in th e droplet breakup on rebound after impingement and reduces the spreading of the droplet as well. The ma ximum recoil height in also reduced with increasing mass concentration. The heat fluxe s of the pure bulk fluids and the ethanol nanofluids did not show any si gnificant enhancement. Th e DMSO nanofluid did show significant enhancement in heat flux when compared to the bulk fluid. Sefiane et al. (2009) researched the evaporation kine tics and wetting dynamics on rough heated surfaces of alumina oxide nanoparticles suspende d in ethanol. The experiment looked at the shape of the droplets by measuring the c ontact angle, base diameter, and volume as a function of time. The pinning of the drops on the heated surface became very important factor. The ethanol with nanoparticles took a longer period of time to pin itself to the solid surface and therefore lead to a decrease in evaporation rate when compared to the base fluid alone. The contact angles for the nanofluid were found to be larger during the depinning process than for the base fluid. The total evaporation time was found to be longer for the base fluid compared to the nanof luid. Again, contrary to the increase in thermal conductivity and heat transfer coeffici ent, the addition of nanoparticles has had an adverse effect on phase change heat transfer. Table 1: Summary Table Enhancing Effects Deteriorating Effects References pH Effects Keep pH level away from isoelectric point Increases the dispersion of nanoparticles, hydration forces and ability for heat Anoop et al. (2009)
20Table 1: Summary Table (Continued) Increase in pH level Thermal conductivity ratio decreases Xie et al. (2002) Transient Hot Wire Method Nanoparticle volume fraction Higher volume fraction results in an increase in thermal conductivity Xuan et al. (2000) Base fluid thermal conductivity Lower thermal conductivity fluids will benefit more from the addition of nanoparticles Fluids with high thermal conductivities will benefit little from the addition of nanoparticles Hwang et al. (2006) Nanoparticle thermal conductivity and thermal diffusivity dependence on temperature Does not have an effect on the thermal conductivity of the nanofluid Zhang et al. (2006) Nanoparticle surface area An optimum specific surface area exist Xie et al. (2002) Pool Boiling Delay of nucleate boiling regime Inconsistent with the increase of thermal conductivity of nanofluids Bang et al. (2005) Fouling of the heated surface by nanoparticles Decrease in nucleation site density Bang et al. (2005) CHF enhancement Not related to the increase in thermal conductivity by the addition of nanoparticles You et al. (2003) Increase in viscosity of the base fluid by the addition of nanoparticles Increase in heat transfer of the base fluid Das et al. (2003) Decrease in nucleation site density Overshadows the increase in heat transfer Das et al. (2003)
21Table 1: Summary Table (Continued) Very small addition of nanoparticles No effect on boiling heat transfer Trisaksri et al. (2009) Deposition of nanoparticles on surface Introduces a thermal resistance Wen et al. (2008) Increase in wettability Driving mechanism for increase in CHF Coursey et al. (2008) Impinging Jet Nanoparticle concentration, flow rate, and nozzle-tosurface distance There exist an optimum to give the best results Nguyen et al. (2009) Jet boiling curves shifted to the right when using nanofluids Indicates a deterioration of boiling heat transfer Liu et al. (2007) Surface became smoother after using nanofluids Decrease in boiling heat transfer Liu et al. (2007) Prevention of vapor blanket formation by the trapping of liquid in the porous layer Increase in the CHF Liu et al. (2007) Spray Cooling Nanoparticle addition reduces evaporation time Enhancement to heat transfer for evaporative cooling Shen (2009) Addition of high aspect ratio microchannels to the copper surface 200% enhancement to single-phase heat transfer and 181% enhancement to two-phase heat transfer Coursey (2007) Longer evaporation time for the base fluid compared to the nanofluid Heat transfer enhancement Sefiane et al. (2009)
22 Chapter 4 Experimental Setup and Procedure 4.1 Nanofluid Preparation In the current study, Al2O3 nanoparticles were chosen because of their widely known thermal properties and ease of dispersion in de-ionized water. Aluminum Oxide mass concentrations of 0.1%, 0.5% and 1% were used for the investigation. The nanoparticles used were made by Nanophase Technologies Corporat ion. The properties of the nanoparticles are: Table 2: Properties of Alum inum Oxide Nanoparticles Purity Avg. Particle Size Specific Surface Area True Density Morphology 99.5+% 45 nm 45 m2/g 3.6 g/cc Spherical The mass of the de-ionized water was determined on a digital scale at which time the desired mass concentration of alumina nanopar ticles was added. Initial dispersion of the mixture was achieved by sonicating the mixtur e for a minimum of 12 hours by the use of an Ultrasonic Cleaner FS140 sonicator. Some evaporation of the nanofluid occurred due to the temperature rise during sonication. To prevent any significant loss of de-ionized water mass, a lid was placed on the contai ner and any change to the nanofluid mass concentration was assumed to be insignificant. To assure proper alumina nanoparticle dispersion during the experiment the pH of the sonicated nanofluid was altered. An Oakton pH 11 handheld pH meter was used to determine the pH level of the nanofluid.
Since pH levels are a function of temperature, the containe r of hot nanofluids was taken from the sonicator and placed in a pool of room temperature water. Once equilibrium was achieved the pH level of the nanofluid was changed with the use of sodium hydroxide (NaOH) and hydrochloric acid (HCl). The pH levels for the different mass concentrations of alumina nanofluids we re determined from the work of Anoop et al. Though the investigation that was referenced on ly dealt with mass concentrations of 1%, 2%, 4%, and 6%, the data was plotted and extrapolated to apply to the current investigation. The result of the regression yielded: y = 0.0239x2 0.456x + 6.892 R2 = 0.992 0 1 2 3 4 5 6 7 01234567 Mass Concentration [%]pH level Figure 1: pH Level vs. Mass Concentration of Alumina Nanofluids The extrapolated data gave pH values of: Table 3: pH Level of Selected Mass Concentrations Mass Concentration pH Level 0.1% 6.8 0.5% 6.7 1.0% 6.5 23
Visual inspection of the nanof luid af ter pH alteration show ed that after 5 days the alumina nanoparticles maintained good dispersion within the water. This was noticed by the cloudiness of the nanofluid, especially noticea ble near the surface of the container. If the nanofluid was clearer near the top of the container it was assumed that the nanoparticles were not very well dispersed. 4.2 Copper Block The copper block was fabricated out of single piece of tellurium copper. Tellurium copper was chosen for this i nvestigation because of its high thermal conductivity and machin ability. A 25.4 mm2 heated surface was fabricated for this investigation. The copper block was designed to provide a 40 .64 mm long extended surface where three K-type, 30 gage thermocouples were inserted 12.7 mm deep at distances of 1 mm, 11 mm, and 21 mm from the spray surface. The base of the copper block was 76.2 x 76.2 x 50.8 mm and had five hol es fabricated where cartridge heaters were inserted. 24Figure 2: Copper Block Design
The OMEGALUX CIR-2013/120V car tridge heaters were 5 0.8 mm long with a 9.525 mm diameter and had a rated wattage of 500 watts. Through prior experimentation, it was found that only four cartridge heaters were needed to conduct th e investigation. The cartridge heaters where inserted at the ends leaving the center hole empty. Due to the high temperatures produced in the copper block an insulation of concrete was molded and placed on the extended surface. Concrete was chosen because of its minimum expansion with temperature rise, cost effectiveness, c ould be easily reproduced in the laboratory and it sufficiently insulated the extended surface fo r the current investigation. An insulated surface was necessary to justif y assumption of a linear temperature profile. To validate the assumption of a linear temp erature profile through the extended surface and a uniform heat flux at the spray surface a COMSOL m odel was developed. The boundaries of the model experienced convective heat transfer at 293 K and a convective heat transfer coefficient of 40 W/m2K. The material properties of the concrete insulation were given by COMSOLs materials database The volumetric heat flux (q ), generated by the cartridge heaters, was found by the following equation: C R A RV V Pq12 2 where PR is the rated wattage of the cartridge heaters, VA is the actual voltage, VR is the rated voltage and C is the circumferential volume of the cartridge heaters. The following figure demonstrates the boundary temp erature profile for the copper block with 15 volts of actual voltage to the cartridge heaters: 25
Figure 3: Boundary Temperature Profile The heat flux path is shown to be linear through the extended surface of the copper block. Figure 4: Heat Flux Path through Block 26
A unifor m heat flux normal to the spray surface is important for accurate calculations during the experiment. The model shows that the insulation adequately provides this uniformity. Hole for thermocouple Figure 5: Heat Flux Normal to Spray Surface 4.3 Spray System The working fluid was poured into a pr essure tank that was pressurized by a compressed nitrogen tank. The flow of the working fluid was regulated by a flow meter connected to a Tefen standard conical spray nozzle. 27
Figure 6: Schematic of Spray System The nozzle was designed to deliver a uniform size and distribution of the droplets. The distance between the nozzle and the heated surface was maintained at 21 mm. The mass flow rates used in this investigation were: Table 4: Mass Flow Rates Pressure [psi] Mass flow rate [g/s] 40 0.53 45 0.58 50 0.61 4.4 Spray Surface Preparation The heated copper surface wa s cleaned after every trial to ensure that the surface characteristics were maintained relatively unchanged from one trial to the next. After spray cooling with both water and the nanofluids, thin film s were observed on the heated surface. A layer of oxidation was caused by the water and a thin film of alumina nanoparticles were deposited by the nanofluid. After the copper block was allowed to 28
29 reach room temperature a liberal amount of Vishay Measurements Group, Inc. M-PREP conditioner was placed on the spray surface and wet-lapped 20 times in the same direction with 320 grit sandpaper to ensure uni formity of the surface. A clean gauze was used to dry the surface after wet-lapping. Fi nally, M-PREP neutraliz er was applied with clean cotton-tipped applicators a nd the surface was dried once again. 4.5 Acquisition System A computer with an acquisition system made by National Instruments was used to acquire data for this investigation. The thermocouples were conn ected to a NI SCXI1303 terminal block. This block is designed specifically for high-accuracy thermocouple measurements and minimizes errors by using an isothermal construction. The data was displayed on the computer by the use of LabVIEW 7.1 software. A program was written that would display the temperature of each thermocouple simultaneously as a function of time.
Figure 7: LabVIEW Front Panel The waveform chart was used to determin e when a steady state condition had been reached. The resolution of the program was 1 sample at a rate of 10 Hz, which gave a good description of the measured transient temperatures data. 4.6 Surface Roughness Measurement To study the effects on the surface by spray cooling with nanofluids, measurements of its surface roughness were ma de. To measure the roughness profile a Surtronic 3P profilometer was used. The profilometer used a diamond tip stylus with a diameter of 5 m. The profilometer was able to compute and display common surface roughness values. The cutoff length of the prof ilometer was 0.8 mm. That meant that the profilometer could not detect any deviation from the normalized data greater than 0.8mm. 30
31 4.7 Experimental Procedure The experimental setup consisted of an open spray system and the copper block was oriented horizontally on a metal sta nd. The effectiveness of different mass concentrations of alumina nanofluids were co mpared to de-ionized water at the same nozzle pressure and distance from the heated surface. The experimental procedure was repeated three times at each concentration and pressure to arrive at an average. The mass of the de-ionized water was measured and th e required alumina nanoparticles were added to achieve the desired mass concentration. Th e mixture was then sonicated for at least 12 hours to disperse the nanopartic les. After sonication, the mi xture was allowed to reach a temperature of 25 C in a cooling bath. The pH of th e nanofluid was altered to maintain the nanoparticles in dispersion for the duration of the experiment. The nanofluid was poured into the pressure tank and the desire d spray nozzle pressure was set by using the compressed nitrogen tank. The thermocouples were inserted into th e extended surface of the copper block and the insulation was placed. The electrical cart ridge heaters were inserted into the copper bloc k base and energized. The fl ow meter was fully opened and the spray cooling of the surf ace began. Once steady state was achieved, the temperatures of the three thermocouples were recorded a nd the voltage to the cartridge heaters was increased gradually until critical heat fl ux (CHF) was reached. After concluding the experiment, the thermocouples, insulation, a nd cartridge heaters were removed and the copper block was allowed to cool. Once cooled, the spray surface was cleaned and prepared for the next experiment.
Chapter 5 Results an d Discussion 5.1 Uncertainty Analysis In the current investigation, the uncertain ties of the heat-fl ux calculations were dependent on the uncertainty of the temper ature readings and the distance between the thermocouples. To measure the uncertainty of the temperature read ings, the uncertainty of the thermocouples and the DAQ (Data Acqui sition) board became important. First, the uncertainty of the thermocouples had to be expressed in terms of a voltage. The sensitivity (STC) of the thermocouple was found by di viding the thermoelectric voltage (VTE) of the thermocouple by the corresponding temperature (T). CT mVV STE TC To find the uncertainty of the thermocouple in terms of voltage (UTC,V), the sensitivity was then multiplied by the uncertainty of the thermocouple (UTC,T) in degrees Celsius, which was 2.2 C. CU C mV SUTTC TC VTC , The uncertainty of the DAQ board (UDAQ) was found by dividing the voltage range (VR) by 2 raised to the resolution of the board, which was 16 bits. 162 mVV UR DAQ 32
W ith the uncertainty of the thermocouple and the DAQ board both in terms of voltages, the voltage uncertainty of the readings (UV) could be found by: 2 2 DAQ VTC VUUU Finally, the uncertainties of the temperature readings (UT) were found by converting the voltage uncertainty (UV) using the scaling function in the LabVIEW software. The scaling function is used by LabVIEW to convert a measured voltage to temperature. The conversion was given by: UT = UV ((2.508355 E -2) + UV ((7.860106E -8) + UV ((-2.503131 E -10) + UV ((8.315270 E -14) + UV ((-1.228034 E -17) + UV ((9.804036 E -22) + UV ((-4.413030 E -26) + UV ((1.057734 E -30) + UV (-1.052755 E -35))))))))) The scaling function has a range of 0 C to 500 C. The distance between the thermocouples was found by a caliper with a re solution of 0.001 meters Therefore, the uncertainty of the distance (UC) was found by taking half the resolution. m m Uc0005.0 2 001.0 The uncertainty of the heat flux (Uq) was found by considering the uncertainties of the temperature readings (UT) and the distance between the thermocouples (UC). ,2 2 2 L U T U qUC T q where q is the calculated heat flux between the thermocouples at 1 mm and 11 mm from the heated surface, T is the temperature difference between the two thermocouples, L is 33
the distance between the therm ocouples, UC is the uncertainty of the distance between the two thermocouples, and U2 T is the temperature uncertainty of the temperature difference between the thermocouples and is given by: 2 11, 2 1, 2 T T TUUU where UT,1 and UT,11 are the temperature uncertainties at distances of 1 mm and 11 mm from the heated surface respectively. The uncertainty analysis revealed that the uncertainty of the heat flux m easurements were approximately 4.6%. 5.2 Experimental Results In this investigation the heat flux rem oved from the heated surface was calculated by using one-dimensional conduction through the extended surface: 11..1 11..1" L T kq where k is the thermal conductivity of the copper block, T1..11 is the temperature difference between thermocouples at distan ces of 1 mm and 11 mm from the heated surface, and L1..11 is the distance between the thermo couples. The heat flux was plotted against the temperature of the surface minus the temperature of the working fluid. The temperature of the working fluid was approximately a constant 23.5 C throughout the length of the experiment. To find the temperature of the surface th e heat flux calculated between the thermocouples at 1 mm and 11 mm from the surface was assumed to be equal to the heat flux between the surface a nd the first thermocouple. Therefore, the surface temperature could be calculated by: 34
11..TL k q TS S where q is the calculated heat flux, k is the thermal conduc tivity of the copper, LS..1 is the distance between the surface and the first thermocouple and T1 is the temperature of the thermocouple at 1 mm from the surface. De-ionized water was first investigated at the different operating pressures. The results of the de-ionized water spray cooling heat transfer curves were compared to investig ate the role of pressure on heat transfer. 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 8: Spray Cooling Curve for Water at 40 Psi 35
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160 Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 9: Spray Cooling Curve for Water at 45 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160 Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 10: Spray Cooling Cu rve for Water at 50 Psi 36
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] 50 psi 40 psi 45 psi Figure 11: Spray Cooling Curve Comparis on of Water at Different Pressures Comparing the spray cooling heat transfer curv es of water at the three different pressures showed that with increasing pressure, the he at transfer at the surface also increased. These results were expected because when th e pressure is increased it results in an increase in the mass flow rate of water drople ts being delivered to the heated surface. The CHF values at the corresponding temperatures are given below: Table 5: Critical Heat Flux for Water Pressure [Psi] Critical Heat Fl ux [W/m^2] Temperature [Celsius] 40 110,833 106 45 119,000 104.8 50 129,500 105.1 The data shows that increasing the pressure results in an increase in the CHF by 7.4% and 8.8% when going from 40 to 45 Psi and 45 to 50 Psi respectively. After the completion of the water data, one of the four cartridge heaters malfunc tioned. The experiments for the nanofluid part of the i nvestigation was done with only three cartridge heaters, one inserted in the center and one on either side. As a result of usi ng only three cartridge 37
heaters, m ore data points were collected during the spray cooling experiments with nanofluids. With only three ca rtridge heaters the heat flux ge nerated at the same variac voltage was insufficient to reach CHF. Ther efore, the number of times the variac was incrementally increased to reach CHF was highe r with three cartridge heaters than with four. The investigation began by l ooking at 1.0% mass concentration of alumina nanofluid. 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 12: Spray Cooling Curve for 1.0% wt. Alumina Nanofluid at 40 Psi 38
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 13: Spray Cooling Curve for 1.0% wt. Alumina Nanofluid at 45 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 14: Spray Cooling Curve for 1.0% wt. Alumina Nanofluid at 50 Psi 39
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] 40 Psi 45 Psi 50 Psi Figure 15: Spray Cooling Curve Co mparison of 1.0% wt. Alumina Nanofluid at Different Pressures Similar to the results obtained for wate r, the heat flux obtai ned by using alumina nanofluids increased with increasing pressu re. The CHF values at the corresponding temperature for each pressure are given below: Table 6: Critical Heat Flux for 1.0% wt. Alumina Nanofluids Pressure [Psi] Critical Heat Fl ux [W/m^2] Temperature [Celsius] 40 133,000 144.7 45 140,000 143.9 50 154,000 143.8 An increase in the CFH of 10% resulted from an increase in pressure from 45 to 50 Psi compared to only a 5.3% increase when increasing the pressure from 40 to 45 Psi. The results for 0.5% wt. concen trations are shown below. 40
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 16: Spray Cooling Curve for 0.5% wt. Alumina Nanofluid at 40 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 17: Spray Cooling Curve for 0.5% wt. Alumina Nanofluid at 45 Psi 41
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 18: Spray Cooling Curve for 0.5% wt. Alumina Nanofluid at 50 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] 40 Psi 45 Psi 50 Psi Figure 19: Spray Cooling Curve Comparison for 0.5% wt. Alumina Nanofluids at Different Pressures 42
As expected the increase in pressure causes an increase in the heat flux removed from the heated surface. The CHF values at th e corresponding temperatures for each pressure are given below: Table 7: Critical Heat Flux for 0.5% wt. Alumina Nanofluid Pressure [Psi] Critical Heat Fl ux [W/m^2] Temperature [Celsius] 40 126,000 145.4 45 129,500 144.7 50 143,500 142.5 Increasing the pressure from 40 to 45 Psi only yielded a 2.8% increase in the CHF for 0.5% wt. alumina nanofluid. A more signifi cant increase of 10.8% was noticed in the CHF when the pressured was raised from 45 to 50 Psi. Finally, the 0.1% wt. concentration results are given below. 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 20: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 40 Psi 43
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 21: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 45 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] Figure 22: Spray Cooling Curve for 0.1% wt. Alumina Nanofluid at 50 Psi 44
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] 40 Psi 45 Psi 50 Psi Figure 23: Spray Cooling Curve Comp arison for 0.1% wt. Alumina Na nofluid at Different Pressures Once again, increasing the pressure resulted in an increase in the heat flux at the spray surface. The CHF data collected and the co rresponding temperature for each pressure is given below: Table 8: Critical Heat Flux for 0.1% wt. Alumina Nanofluid Pressure [Psi] Critical Heat Fl ux [W/m^2] Temperature [Celsius] 40 115,500 145.2 45 122,500 144.7 50 133,000 142.3 Increasing the pressure from 40 to 45 Psi resu lts in an increase of 6.1% to the CHF and increasing the pressure from 45 to 50 Psi gives an 8.6% increase. The spray cooling experiments show the same results for water and alumina nanofluids, increasing the mass flow rate of droplets enhances heat transfer at the surface. The objective of the study was to investigate enhancements when compared to water at the same pressure. Therefore, the alumina nanofluid data was compared to water at the same pressure. 45
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] water 0.1% wt. 0.5% wt. 1.0% wt. Figure 24: Spray Cooling Curve Comparison of Water vs. Nanofluids at 40 Psi 0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] water 0.1% wt. 0.5% wt. 1.0% wt. Figure 25: Spray Cooling Curve Comparison of Water vs. Nanofluids at 45 Psi 46
0 20000 40000 60000 80000 100000 120000 140000 160000 0102030405060708090100110120130140150160Ts-Tf [Celcius]Heat Flux [W/m^2] water 0.1% wt. 0.5% wt. 1.0% wt. Figure 26: Spray Cooling Curve Comparison of Water vs. Nanofluids at 50 Psi The data shows that the addition of alumin a nanoparticles to water had a positive effect on single-phase and part of tw o-phase heat transfer during spray cooling experiments. The data also shows a shift to the right of the spray cooling curve, indicating a delay in two-phase heat transfer for all three pre ssures investigated. The heat transfer enhancement can be seen by an upward shif t of the spray cooling curve when using alumina nanofluids. For example, at a pressu re of 50 Psi and a temperature difference of approximately 79 C, the heat flux at the spray surface for 1.0% wt. alumina nanofluid is calculated as 63,000 W/m2 compared to only 44,333.3 W/m2 with water. That result, shows a 42% increase in the heat flux remove d from the heated surface. One possible explanation for the enhancement in heat tr ansfer at the surfa ce is the increase in wettability of the water by the addition of nanopa rticles. Wetting is th e ability of a liquid to remain in contact with a solid surface. Coursy et al. (2007) cited the increase in wettability as a possible mechanism in his pool boiling experiments. Since, the copper 47
48 spray surface was oriented horizontally the dr oplets traveled across th e heated surface, by the force of gravity, removing heat. If the waters wettability increased with the addition of alumina nanoparticles, the droplets surface area in contact with the surface increased as they moved along the surface, therefore increasing heat transfer at the surface. Another mechanism for the increase in single-phase heat transfer is the time it takes for a droplet to travel the length of the h eated surface. The increase in wettability will make the droplets attach to the surface l onger increasing the ability for the droplet to remove heat. The data also shows that the mass concentratio ns of nanoparticles have little effect on the heat transfer enhancement during spray cooling. The nanofluids also showed enhancements to the CHF at all three pressu res. The CHF enhancement was noticed to be effected by the mass concentrations of the nanofluids. At a mass concentration of 1.0% wt. the CHF had an average increase of 18.8%. An average increase of 11.1% and 3.3% was achieved with 0.5% wt. and 0.1% wt. mass concentrations respectively. The spray cooling experiments with nanofluids also showed a delay in two-phase heat transfer. The delay is characte rized by a shift to the right of the spray cooling curve. One possible mechanism investigated for the increase in CHF and the delay in two-phase heat transfer was the surface roughness of the spray surface. The nanoparticles used in this investigation were a number of magnitude s smaller than the su rface roughness of the spray surface. The nanopart icles are deposited to the su rface by the vaporized water droplets. As a result, the nanoparticles b ecome impinged in the surface crevices and change the characteristics of the surface. Once a layer of na noparticles is deposited onto the surface, a new thermal resistance is intr oduced and the number of nucleation sites is reduced. The heat flux at the surface will have to be conducted through the deposited
alum ina nanoparticles, which have a lower th ermal conductivity than the copper surface, before being removed by the spray cooling pr ocess. A profilometer was used to measure the surface roughness of the spra y surface before and after sp ray cooling with nanofluids and after the cleaning procedur e had been performed. -4 -3 -2 -1 0 1 2 3 4 0.42.44.46.48.410.412.414.4Distance [mm]Z value [ m] Ra = 1.15 m Figure 27: Surface Roughness before Spray Cooling 49
-4 -3 -2 -1 0 1 2 3 4 0.42.44.46.48.410.412.414.4Distance [mm]Z value [ m] Ra = 0.89 m Figure 28: Surface Roughness after Spray Co oling with 0.5% wt. Alumina Nanofluid -4 -3 -2 -1 0 1 2 3 4 0.42.44.46.48.410.412.414.4Distance [mm]Z value [ m] Ra = 1.06 m Figure 29: Surface Roughness after Cleaning Procedure The results of the surface roughness measurem ents show the effects by the addition of alumina nanoparticles to water. The averag e roughness (Ra) value before spray cooling is found to be 1.15 m. After spray cooling with a ma ss concentration of 0.5% wt. the surface roughness is measured again a nd found to have decreased to 0.89 m. The 50
51 results of the surface roughness measuremen ts indicate that the impinged alumina nanoparticles have made the copper surface smooth er. To ensure the repeatability of the experiment, the surface roughness was measur ed after the cleaning procedure was performed. The cleaning pro cedure returned most of the roughness back to the surface and was found to be 1.06 m. The impinged alumina nanoparticles on the copper spray surface have decreased the nucleation site density of the surface where the droplets change phase into vapor form. The reduc tion of vapor on the heated surface caused a delay in two-phase heat transfer. Two-phase heat transfer is desirable because it is a more effective way to remove heat when compared to single-phase heat transfer. Twophase heat transfer utilizes the latent heat of evaporation of the wo rking fluid to cause a phase change from liquid to vapor. This process is endothermic, which means that energy is absorbed by the droplets from the heat ed surface in going from liquid to vapor. Since a vapor blanket cannot form as easil y once the surface has become fouled by the alumina nanoparticles, an increase in the CH F during spray cooling is found to occur. During pool boiling experiments CHF is characte rized by a layer of vapor that forms at the heated surface preventing the working fluid from coming in contact with the surface, resulting in an increase in te mperature. Similarly, during the spray cooling experiments, a vapor blanket formed over the heated copper surface which prevented the droplets from impinging the surface. The hot vapor blanke t over the surface is not effective at conducting heat away from the surface, because of the low heat transfer coefficient of the vapor, which results in an increase in the temperature of the spray surface. The delay in two-phase heat transfer cause d by the impingement of alumina nanoparticles allows for heat transfer to continue pa st the CHF point of water. The higher surface temperatures
52 experienced during the delay increased the heat flux at the surface and led to an increase of the CHF when alumina nanofluids where used as the working fluid. Higher CHF values resulted when using higher mass concen trations of alumina nanoparticles, though a further delay in CHF was not a function of mass concentration. Theoretically, the higher mass concentration alumina nanoflu ids deposit more nanoparticles onto the surface than the lower concentrations during the length of the experiment. This could have lead to less vapor a nd higher temperatures with higher mass concentrations.
53 Chapter 6 Conclusion and Recommendations 6.1 Conclusion The results of the investigation show th at adding nanoparticles to the de-ionized water enhanced single-phase he at transfer as indicated by an increase in heat flux at the surface by as much as 42% when compared to water at the same temperature difference and pressure. One reason for this enhan cement could be the change of the hydrodynamic characteristics of water. Th e addition of nanoparticles made the water more wettable and increased the wetting angle of the droplets. The droplets were able to remain in contact with the heated copper surface l onger, increasing their effectiv eness to remove heat. The horizontal position of the heated surface had an effect on the enhancement as well. With the horizontal orientation the droplets that impinged the surface at the top of the heated surface dragged across the surface by the for ce of gravity and heat was removed more effectively. The mass concentration of the nanopa rticles seemed to have little to no effect on the single-phase heat transf er enhancement but did show effects with the increase in the CHF. All concentrations of nanoparticles resulted in a delay of two-phase heat transfer during the spray cooli ng investigation. The decrease in nucleation site density delayed the formation of vapor and increased th e thermal resistance at the spray surface. The delay of two-phase heat transfer created higher surface temperatures which led to the increase in CHF. The higher mass concen tration of 1.0% wt. re sulted in an average
54 increase of 18.8% when compared to 0.5% wt and 0.1% wt. with increases in CHF of 11.1% and 3.3% respectively 6.2 Recommendations For future studies it will be important to investigate the results of altering the pH level of the nanofluid, since it has effect s on the thermophysical properties of the nanofluid. The effects on the hydrodynamic properties of water by the addition of nanoparticles should also be considere d. These properties could explain the enhancements to single-phase heat transfer and CHF. The orientation of the heated surface should be changed and its effects investigated. The copper block design could be improved to provide better efficiency of deliv ering the heat flux to the heated surface and not loosing much of it to the environment through the insulation. Much lower mass concentrations of nanoparticles, in the order of 0.001%, should be investigated to find an optimum concentration. To decrease th e amount of nanofluids used during the investigation, a closed-loope d system should be used.
55 References Anoop, K. B., T. Sundararajan, and Sarit K. Das. "Effect of Particle Size on the Convective Heat Transfer in Nanof luid in the Developing Region." International Journal of Heat and Mass Transfer 52 (2009): 2189-95. Print. Bang, Cheol In, and Soon Heung Chang. "Boili ng Heat Transfer Performance and Phenomena of Al2O3water Nano-Fluids from a Plain Surface in a Pool." International Journal of Heat and Mass Transfer 48 (2005): 2407-19. Web. Choi, Stephen U. S., and J. A. Eastman. "Enhancing Thermal Conductivity of Fluids with Nanoparticles". ASME International Mechanical Engineering Congress and Exposition. November 12-17, 1995, San Francisco, CA. 1995. Print. Chon, Chan Hee, Kenneth D. Kihm, and Shin Pyo Lee. "Empirical Correlation Finding the Role of Temperature and Particle Size for Nanofluid (Al[Sub 2]O[Sub 3]) Thermal Conductivity Enhancement." Applied Physics Letters 87 (2005): 153107. Print. Coursey, Johnathan S., and Jungho Kim. "Na nofluid Boiling: The Effect of Surface Wettability." International Journal of Heat and Fluid Flow 29.6 (2008): 1577-85. Print. Coursey, Johnathan S. "Enhancement of Spra y Cooling Heat Transfer using Extended Surfaces and Nanofluids." Ph.D. Univ ersity of Maryland, 2007. Print. Das, Sarit K., Nandy Putra, and Wilfred Roet zel. "Pool Boiling Characteristics of NanoFluids." International Journal of Heat and Mass Transfer 46 (2003): 851-62. Print. Duursma, Gail, Khellil Sefiane, and Aiden Kennedy. "Experimental Studies of Nanofluid Droplets in Spray Cooling." Heat Transfer Engineering 30.13 (2009): 1108-20. Print. Hwang, Y. J., et al. "Investigation on Characteristics of Thermal Conductivity Enhancement of Nanofluids." Current Applied Physics 6 (2006): 1068-71. Print. Incropera, Frank P., et al. "Impinging Jets." Introduction to Heat Transfer. 5th ed.Wiley, 2006. 402. Print.
56 Jang, Seok Pil, and Stephen U. S. Choi. "R ole of Brownian Motion in the Enhanced Thermal Conductivity of Nanofluids." Applied Physics Letters 84 (2004): 21-4. Print. Liu, Zhen-Hua, and Yu-Hao Qui. "Boiling Heat Transfer Characteristics of Nanofluids Jet Impingement on a Plate Surface." International Journal of Heat and Mass Transfer 43 (2007): 699-706. Print. Murshed, S. M. S., K. C. Leong, and C. Yang. "Enhanced Thermal Conductivity of TiO2water Based Nanofluids." International Journal of Thermal Sciences 44.4 (2005): 367-73. Print. Nguyen, Cong Tam, et al. "An Experimental Study of a Confined and Submerged Impinging Jet Heat Transfer us ing Al2O3-Water Nanofluid." International Journal of Thermal Sciences 48.2 (2009): 401-11. Print. Sefiane, K., and R. Bennacer. "Nanofluid s Droplets Evaporation Kinetics and Wetting Dynamics on Rough Heated Substrates." Advances in Colloid and Interface Science 147 (2009): 263-71. Print. Shen, Jian. "Hydrodynamics of Droplet Impi ngement on Heated Surfaces : Effects of Nanofluid and Nano-Structured Surface." Ph.D. Oregon State University, 2009. Web. 10/12/2009. Trisaksri, Visinee, and Somchai Wongwises. "Nucleate Pool Boiling Heat Transfer of TiO2-R141b Nanofluids." International Journal of Heat and Mass Transfer 52 (2009): 1582-8. Print. Vassallo, Peter, Ranganathan Kumar, and Ste phen DAmico. "Pool Boiling Heat Transfer Experiments in silicawater Nano-Fluids." International Journal of Heat and Mass Transfer 47.2 (2004): 407-11. Print. Wen, Dongsheng. "Mechanisms of Thermal Nanof luids on Enhanced Critical Heat Flux (CHF)." International Journal of Heat and Mass Transfer 51.19-20 (2008): 4958-65. Print. Xie, Huaqin, et al. "Thermal Conductivit y Enhancement of Suspensions Containing Nanosized Alumina Particles." Journal of Applied Physics 91.7 (2002): 4568-72. Print. Xuan, Yimin, and Qiang Li. "Heat Tran sfer Enhancement of Nanofluids." International Journal of Heat and Fluid Flow 21.1 (2000): 58-64. Print.
57 You, S. M., J. H. Kim, and K. H. Kim. "Eff ect of Nanoparticles on Cr itical Heat Flux of Water in Pool Boiling Heat Transfer." Applied Physics Letters 83.16 (2003): 3374-6. Print. Zhang, Xing, and Hua Gu. "Effective Thermal Conductivity and Thermal Diffusivity of Nanofluids Containing Spherical and Cylindrical Nanoparticles." Journal of Applied Physics 100.4 (2006): 044325. Print. Zhou, D. W. "Heat Transfer Enhancemen t of Copper Nanofluid with Acoustic Cavitation." International Journal of Heat and Mass Transfer 47.14-16 (2004): 3109-17. Print. Zhu, Dongsheng, et al. "Dispersion Behavior and Thermal Conductivity Characteristics of Al2O3H2O Nanofluids." Current Applied Physics 9.1 (2009): 131-9. Print.