Effect of Nanofluids on Liquid-solid Heat Transfer on High-temperature Wall

Abstract

The industrial application of nanofluids had been explored by many researchers since nanofluids were proposed. However, there were different opinions on the effect in jet cooling. In this paper, 0.4 vol%, 0.8 vol%, 1.2 vol%, 1.8 vol%, 2.4 vol% Al₂O₃-water, TiO₂-water, SiO₂-water nanofluids and pure water were used as quenching coolants to complete single jet cooling experiments on the free surface of 50 mm high-temperature steel plate. The results showed that using low concentration (0.4–1.2 vol%) nanofluids could significantly improve the maximum heat fluxes, cooling speed peaks, and moving velocities of peaks along the thickness direction compared with pure water. However, the cooling uniformity in the horizontal direction was reduced, especially with high concentration nanofluids (≥1.8 vol%). Through comprehensive comparison, when 1.2 vol% Al₂O₃ + water was used as coolant, the optimal cooling efficiency could be achieved, and cooling speed peaks along the thickness were 8.14%–19.70%, 2.16%–3.48% and 0.74%–1.44% higher than that of pure water respectively.

1. Introduction

Quenching can be defined as a heat transfer process in which high temperature solid contacts low temperature fluid instantaneously, resulting in an extremely fast cooling. When the surface of plate is in direct contact with cooling water, the heat can be effectively transferred, while when the high temperature wall is covered by water vapor, the cooling efficiency is significantly reduced.¹⁾ During jet cooling, high-pressure medium impacts the surface of the steel plate to quickly destroy vapor film formed by the boiling cooling medium. Compared with immersion and spray cooling, jet cooling can make plate surface reach a higher heat flux, and high-strength steel with better mechanical properties is finally obtained. When the jet impinges on the high temperature wall, a dark black wetted area appears at the jet point. As cooling progresses, this area gradually expands around the jet point. The visible outer edge can be called wetting front, and the transition boiling zone is considered as a part of wetting front.²⁾ Filipovic et al.³⁾ studied the wetting front propagating along the flow direction. They believed that there was nuclear boiling or single phase forced convection upstream of the wetting front, and film boiling downstream. The position of the surface maximum heat flux moved downstream with cooling. Mozumder et al.²⁾ considered that the starting position of nuclear boiling was a little upstream of intense boiling zone, and MHF was at the boundary of nuclear boiling and transition boiling. The position of the MHF usually corresponded to the maximum temperature gradient. How to control the MHF during quenching was an important goal to study solid-fluid heat transfer of high temperature wall. Many researchers had used MHF as one of the evaluation criteria for cooling capacity to study and optimize the parameters such as cooling temperature,⁴⁾ jet flow,⁵⁾ jet height,⁶⁾ jet angle,⁷⁾ nozzle shape⁸⁾ and nozzle distribution.⁹⁾ Some of the results had been used in industrial production, but the effect of improving cooling strength and uniformity was not very obvious. The problems of low production efficiency, increased energy consumption, and limited mechanical properties of steel still existed in the metal plate heat treatment field.

Improving the heat transfer performance of the cooling medium to heighten the thermal efficiency had become an important method to enhance the fluid-solid heat transfer of high-temperature walls.¹⁰⁾ In recent years, with the rapid development of nanomaterial research, the preparation technology had become more perfect, which made the application of nanoparticles in the steel material heat treatment field became feasible. Nanofluid was a stable, uniform, new heat transfer cooling medium formed by dispersing nanoparticles such as metal or non-metal into a base fluid such as water, alcohol or oil by a certain method and ratio, which was first proposed Choi et al of the Argonne National Laboratory in the United States in 1995.¹¹⁾ Due to the larger surface to volume ratio, fluid with nanoparticle had a higher thermal conductivity than fluid with microparticles or pure fluid.¹²⁾ Many researchers had studied the effect of nanofluids on thermal conductivity. Murshed et al.¹³⁾ used Φ10 nm × 40 nm and Φ15 nm TiO₂ particles with a volume fraction of up to 5%. The study found that addition of nanoparticles increased thermal conductivity of the base fluid by 33% and 30%, respectively. Aberoumand et al.¹⁴⁾ added 1.0 vol% CuO to the oil, and the results showed that the thermal conductivity increased by 49% compared with the base fluid. The increase in thermal conductivity would improve the convective heat transfer capacity of the fluid. At present, the combination of nanofluid and jet was considered to be an effective technique to enhance heat transfer. Yousefi et al.¹⁵⁾ experimentally studied the vertical jet on the apex of V-plate when the nanofluid volume fraction was 0.05% and found that the maximum heat transfer coefficient increased by 13.91%. Jizu et al.¹⁶⁾ completed a free single jet impact experiment using Al₂O₃-water nanofluid as a coolant. The results show that the heat transfer coefficient increased by 61.4% when the nanoparticle volume fraction was 2.0%. using Cu-water nanofluid, Li et al.¹⁷⁾ increased heat transfer rate of a single jet by 52%. Nguyen et al.¹⁸⁾ had studied restricted jet and results showed that 2.8 vol% Al₂O₃-water nanofluids could obtain the strongest cooling capacity when the jet height was 5 mm, while 6% and higher volume fractions were not suitable for improving heat transfer performance. The addition of nanoparticles resulted in a significant increase in the dynamic viscosity compared with base fluid.^19,20,21,22) Sidi et al.²³⁾ thought that nanoparticles had a dramatic effect on wall shear stress, and the shear stress increased with the increase of particles. Alkasmoul et al.²⁴⁾ considered that the negative impact of nanofluids on kinematic viscosity exceeded all gain in thermal conductivity. When using nanofluids, achieving a target temperature at a given thermal load needed larger flow rate and more pump power, and nanofluids could not provide any practical benefit. According to the statistics of Sergis et al.,²⁵⁾ 19% of research papers showed that the use of nanofluids could improve convective heat transfer effect by 10% to 18%, while 11% of papers showed deterioration. Whether nanofluids could replace base fluids as coolants required further research.

At present, many researchers still had different opinions on whether nanofluids had beneficial effects on fluid-solid heat transfer. In addition, there were few studies on heat transfer of nanofluids on the high-temperature wall of steel plate. The application of nanofluids as coolants in the field of high temperature metal plate heat treatment required further studying. In this paper, single jet experiments on free surface of 50 mm thickness high-temperature steel plate was completed to study the effect of nanofluids as coolants on the heat transfer performance of high temperature wall. The best nanofluid coolant formulation would be selected based on the experimental results. In this study, 0.4 vol%, 0.8 vol%, 1.2 vol%, 1.8 vol%, 2.4 vol% Al₂O₃-water, TiO₂-water, SiO₂-water nanofluids were used as coolants, and compared with pure water. The heat transfer effect of nanofluids was studied from the aspects of nanofluid species and volume concentration. The research results would obtain the fluid-solid heat transfer law of nanofluids as coolants, and determine the optimal nanofluid solution ratio to improve the wall heat transfer. The study would provide a method and data support for the application of nanofluid in the field of metal sheet strip heat treatment.

2. Nanofluid Single Jet Experiment System and Process

2.1. Experimental Material

The experimental material was 50 mm (thickness) × 80 mm (width) × 150 mm (length) AISI304 austenitic stainless steel. The change of specific heat c_s and thermal conductivity λ_s with temperature were measured by DSC-404-C differential scanning calorimeter and LFA-447 laser thermal conductivity meter. The fitting formulas were shown in Eqs. (1) and (2). The density ρ_s was constant as 7930 kg·m⁻³. The positions of temperature measurement points were shown in Fig. 1. A total of 15 points were designed to measure cooling temperatures at different heat exchange zones on the wall and different positions in the thickness. Holes with a diameter of 3.2 mm and a depth of 40 mm were made at the measurement points for installing thermocouples. K-type armored thermocouples were used to measure the instantaneous temperatures inside the steel plate. The thermocouples model was WRNK-171, and specification was Φ3 × 1000 × 2000 mm. After thermocouples were inserted into the temperature measurement holes, they were fixed and sealed with high temperature resistant glue. The data acquisition device was a multi-channel temperature collector, and acquisition interval was 1 ms.   

$$c_s=\left(-0.2274\times {10}^{-\frac{T}{788}}+0.62\right)\times 1\mathrm{\hspace{0.17em} }000$$(1)

  

$${\lambda }_s=-31.1\times {10}^{-\frac{T}{1\mathrm{\hspace{0.17em} }703}}+47.7$$(2)

Fig. 1.

Distribution of temperature measurement points.

The two-step method was chosen to prepare Al₂O₃-water, TiO₂-water and SiO₂-water nanofluids. The nanoparticles were from Aladdin McLean, and the average particle sizes of αAl₂O₃, TiO₂, and SiO₂ were 30 nm, 40 nm, and 30 nm. Disperse them into 10 L water by magnetic and ultrasonic stirring to prepare nanofluids having a volume concentration of 0.4 vol%, 0.8 vol%, 1.2 vol%, 1.8 vol%, and 2.4 vol%, respectively. The nanoparticles were uniformly and stably dispersed throughout the water tank, and two-phase fluid formula could be used to estimate the thermophysical properties of nanofluids. In the equations for calculating thermal conductivity k²⁶⁾ and viscosity μ,²⁷⁾ the subscripts np, bf, and nf denoted nanoparticle, base fluid, and nanofluid, respectively, and φ was the particle volume fraction:   

$${\mu }_{nf}={\mu }_{bf}\left(1+2.5\phi \right)$$(3)

  

$$k_{nf}=\frac{2k_{bf}+k_{np}+2\phi \left(k_{np}-k_{bf}\right)}{2k_{bf}+k_{np}-\phi \left(k_{np}-k_{bf}\right)}{k}_{bf}$$(4)

2.2. Experimental System

The jet experiment system was a closed fluid circuit including a 10 L open water tank, 3 mm diameter circular nozzle, pump, valve, flowmeter, etc., as shown in Fig. 2. Nanofluid materials and volume concentrations were recorded in Table 1 and the nanofluids were stored in the tank with temperatures at 26°C ± 0.2°C. The square tubes and double top wire cross clamps were used to fix the circular nozzle at a position of 150 mm ± 2 mm above the experimental plate, and the jet point was located on the surface above the temperature measurement point 1-1. The jet flow was controlled by a valve, and formed a closed-loop control with the measured flow by the flowmeter feedback. The flow was adjusted to 210 L/h ± 3 L/h before experiment. When fluid jetted to the surface, it flowed out from sides of plate and was collected by a recovery box. The fluid flowed back into water tank after oxide scale impurity was removed through filter. The temperature acquisition system was mainly composed of embedded thermocouples, multi-channel temperature collector, twisted pair, and computer. During cooling, thermocouples with error of 1% collected the plate temperatures at measurement points. The multi-channel temperature collector recorded data at a frequency of 1000 Hz. Twisted pair was used for data transmission between collector and computer.

Fig. 2.

Layout of experimental equipment.

Table 1. Nanofluid materials and volume concentrations.

Nanofluid material	Volume concentration (vol%)
Pure water	0.0
Al2O3 + water	0.4/0.8/1.2/1.8/2.4
TiO2 + water	0.4/0.8/1.2/1.8/2.4
SiO2 + water	0.4/0.8/1.2/1.8/2.4

2.3. Data Processing

During cooling, the experimental plate had a significant temperature gradient in the thickness direction. It was necessary to calculate the surface temperatures and heat fluxes at the contact interface by inverse heat transfer.²⁸⁾

Establish thermal differential equation:   

$$\frac{\partial T}{\partial t}=a\frac{{\partial }^{2}T}{\partial x^2},\left(0<x<H,t>0\right)$$(5)

Where H was the thickness of plate (mm); t was the time (s); T was the temperature (°C); a was the thermal diffusivity (m²·s⁻¹): $a=\frac{{\lambda }_s}{{\rho }_s{c}_s}$, λ_s, ρ_s and c_s are the thermal conductivity, density, specific heat of experimental plate, respectively. Inverse heat transfer was used to estimate the boundary heat flux through thermal evolution inside the experimental plate. Use the third boundary condition:   

$$\{\begin{array}{l}-\lambda {\displaystyle \frac{\partial T\left(x,t\right)}{\partial x}}{\text{|}}_{\text{x=0}}=0,\left(x=0,t>0\right)\\ -\lambda {\displaystyle \frac{\partial T\left(x,t\right)}{\partial x}}{\text{|}}_{\text{x=H}}=\{\begin{array}{l}{q}_M=const,\left(t_{M-1}<t<t_M\right)\\ q\left(t\right),\left(t>t_M\right)\end{array}\end{array}$$(6)

Discrete the thickness of plate and cooling time using grids. The temperature measurement points were located at the nodes. Finite difference method was used to solve the instantaneous temperature field:   

$$\{\begin{array}{l}{T}_i^{n+1}=2Fox{T}_{i+1}^n+\left(1-2Fox\right)T_i^n,\left(i=0\right)\\ T_i^{n+1}=T_i^n+Fox\left(T_{i+1}^n-2T_i^n+T_{i-1}^n\right),\left(i<0<I\right)\\ T_i^{n+1}=T_i^n+2Fox\left(T_{i-1}^n-T_i^n\right)-{\displaystyle \frac{2qdt}{\rho cdx}},\left(i=I\right)\end{array}$$(7)

Here, n was the time node (0 < n < t); i was the thickness direction node (0 < i < I, I = H/dx), Fox was Fourier number: $Fox=\frac{adt}{d{x}^2}$.

The sensitivity coefficient was defined as the first partial derivative of temperature to heat flux:²⁹⁾   

$$X\equiv \frac{\partial T\left(x,t\right)}{\partial q}$$(8)

  

$$\{\begin{array}{l}-\lambda {\displaystyle \frac{\partial X\left(x,t\right)}{\partial x}}{\text{|}}_{\text{x=0}}=0,\left(x=0,t>0\right)\\ -\lambda {\displaystyle \frac{\partial X\left(x,t\right)}{\partial x}}{\text{|}}_{\text{x=H}}=\{\begin{array}{l}1,\left(t_{M-1}<t<t_M\right)\\ 0,\left(t>t_M\right)\end{array}\end{array}$$(9)

Establish the sensitivity coefficient field as follows   

$$\{\begin{array}{l}{X}_i^{n+1}=2Fox{X}_{i+1}^n+\left(1-2Fox\right)X_i^n,\left(i=0\right)\\ X_i^{n+1}=X_i^n+Fox\left(X_{i+1}^n-2X_i^n+X_{i-1}^n\right),\left(i<0<I\right)\\ X_i^{n+1}=X_i^n+2Fox\left(X_{i-1}^n-X_i^n\right)-{\displaystyle \frac{2dt}{\rho cdx}},\left(i=I\right)\end{array}$$(10)

Use Taylor expansion and least square technology to establish inverse problems:   

$$q_M=q^{*}+\frac{X_M\left(Y_{K,M}-T_{K,M}\right)}{X_K^2}$$(11)

Here,Y_k,M was the temperature of K measurement point at M time. By iteration calculation, when the relative error of two calculation values of heat flux was less than 0.1%, the iteration ended, and the calculation value was regarded as the heat flux in this period. Dou et al.²⁹⁾ showed that relative error between calculated value and measured value was less than 0.1% using this modeling method.

3. Result and Discussion

3.1. Effect of Nanofluid on Maximum Heat Flux

Based on the measured instantaneous temperatures at a distance of 3 mm from surface, the heat fluxes were estimated with r=0–60 mm (r was the radial distance from jet point). Maximum heat flux (MHF) of 0.4–2.4 vol% Al₂O₃/SiO₂/TiO₂ + water and pure water as coolants were shown in Fig. 3. When the nanoparticle concentrations were 0.4–1.2 vol%, the MHF in different radial positions were almost higher than that reached when pure water was used. This could be attributed to local turbulence and damage to the flow boundary layer caused by Brownian motion of the nanoparticles.³⁰⁾ Using three different nanomaterials, MHF almost reached a maximum at 1.2 vol%. It indicated that in this experimental environment, the volume concentration to the maximum MHF might be independent of the type of nanoparticles. Comparing the MHF of three nanofluids with 1.2 vol%, when Al₂O₃ was used as additives, MHFs were 2.78, 2.45, 2.13, 2.29, and 1.69 MW/m² along the radial direction, which were higher than the other two materials under the same experimental parameters. And MHF obtained using TiO₂ were higher than that using SiO₂. This regular in MHF was clearly consistent with the thermal conductivity of nanoparticle materials.

Fig. 3.

MHF of different nanofluids as coolants. (Online version in color.)

When the concentrations of Al₂O₃ were 1.8 vol% and 2.4 vol%, MHF at r=0 mm were 0.36 and 0.21 MW/m² higher than that of pure water, respectively. When the MHF moved to r=15 mm, the value decreased to 0.08 and 0.02 MW/m². However, when the MHF moved further to r=30 mm, the values were 0.12 and 0.04 MW/m² lower than that of pure water. This phenomenon also occurred when 1.8/2.4 vol% TiO₂/SiO2 + water was used. Nguyen et al.¹⁸⁾ used 0.0–6.0% Al₂O₃-water to study the restricted and immersive jet and found that the optimal concentration when the jet height was 5 mm was 2.8%, which was higher than that of 6.0%. They attributed this phenomenon to an increase in fluid viscosity. However, in this paper, the cooling efficiency was high at the near jet point and poor at the far, the change of MHF could not explained by viscosity. At the near jet point, collision of nanoparticles in the fluid accelerated rupture of closed gas film, and more coolant contacted the high-temperature wall. As the vapor evaporated, nanoparticles were concentrated at the bottom of the bubbles, forming a micro layer.³¹⁾ When the micro layer evaporated, nanoparticles adhered to the surface of steel plate. Small nanoparticles would increase the heat transfer area, which provided more sites for boiling bubbles and accelerated heat transfer. However, with the flow along the radial direction, the velocity decreased gradually. More nanoparticles covered on the surface due to natural convection and precipitation, forming a nano coating. It hindered the contact between fluid and wall, which made the cooling capacity of high concentration nanofluids (≥ 1.8 vol%) lower than that of low concentration nanofluids (0.0–1.2 vol%). Tiara et al.³¹⁾ found that the deposition of metal oxides with a thickness of more than 10 ppm on the surface would increase the thermal resistance and reduce the heat transfer.

Figure 4 showed the surface boiling pictures recorded by camera when the wetting front moved to r=15/30/45/60 mm. The heat of wetted zone was transferred to coolant by jet impingement, and no obvious splashing phenomenon was observed. It showed that there was no obvious boiling in this area. The obvious fluid-gas disturbance could be observed in the wetting front, indicating that intense boiling was taking place here. As the fluid moved along radial direction, subcooling and momentum of fluid decreased gradually. As a result, boiling bubbles developed fully. The ability of coolant to overcome the deflection from the bubbles was reduced. The fluid passed through the steam film and away from surface of steel plate. When r=45 mm, the flow velocity reduced to a very small value. The accumulation of cooling fluid caused a large number of boiling bubbles to appear, which accelerated the cooling. A higher MHF appeared at r=45 mm, and the deflection of coolant became more apparent. When the wetting front moved from r=15 mm to r=60 mm, it could be observed that width of wetting front increased from 1.29 mm to 4.28 mm, 6.86 mm and 17.14 mm. Akmal et al.³²⁾ also found the phenomenon that the width increased gradually with movement. They thought that the decrease of jet velocity and subcooling would increase the width, and driving force of wetting front movement might come from radial temperature gradient.

Fig. 4.

Surface boiling pictures. (Online version in color.)

3.2. Effect of Nanofluid on Cooling Speed Peak

Using 0.4–2.4 vol% Al₂O₃, TiO₂, and SiO₂ + water as coolants, the cooling speeds of 3 mm, 25 mm, and 50 mm from the surface of steel plate were calculated according to the measured temperatures. The cooling speed peaks (v_p) were given in Fig. 5. The v_p at 3 mm from surface had the same distribution pattern as MHF, that v_p first increased and then decreased as the increase of nanoparticle volume concentration. When 1.2 vol% Al₂O₃ + water was used, the v_p at 3 mm were 89.40°C/s, 79.07°C/s, 63.60°C/s, 69.31°C/s, and 47.72°C/s. It was the best cooling efficiency compared with other coolants. This result was consistent with the correlation between the maximum heat flux and cooling speed proposed by Mozumder et al.²⁾

Fig. 5.

Cooling speed peaks of different nanofluids as coolants. (Online version in color.)

When v_p moved from surface to inside of steel plate along the thickness direction, nanofluids with high volume concentrations (≥1.8 vol%) tended to reach a greater cooling speed at a distance of 25–47 mm from the surface. For 1.8 vol% Al₂O₃, v_p increased by 0.32%, 6.44%, 4.18%, 2.68% and 6.24% compared with pure water at the 25 mm; and increased by 1.16%, 5.83%, 3.45%, 0.70% and 4.95% at the 47 mm. A similar phenomenon was observed for 2.4 vol% TiO₂ and 1.8% vol% SiO₂. Combined with the MHF’s change in Fig. 3, that when the nanoparticle concentration was ≥1.8 vol%, MHF was higher near jet point and lower away from the point than other nanofluids, it could be inferred that the cooling ability near jet point was able to inherit to a deeper position from the surface, and the cooling ability away from jet point was weakened by side heat radiation during genetic process in the thickness direction.

When 1.2 vol% Al₂O₃ + water was used as coolant, the cooling speed peaks at different thicknesses were mostly higher than those of pure water (Fig. 6(a)). The v_p increased by 16.43%, 18.86%, 17.10%, 19.70%, 8.14% at 3 mm, 3.48%, 2.42%, 3.02%, 2.51%, 2.16% at 25 mm, and 1.01%, 0.85%, 1.44%, 0.95%, 0.74% at 47 mm compared with pure water. v_p increase rate decreased gradually from the surface to the inside, which could be attributed to the reduced temperature gradient along the thickness.³³⁾ The v_p moved in the thickness direction, and average moving velocities were given in Fig. 6(b). Using 1.2 vol% Al₂O₃ + water, the velocities increased by 6.95%, 0.89%, 5.43%, 0.51%, 9.15% along axial at 0–3 mm from surface, 6.38%, 3.56%, 24.98%, −1.06%, 8.41% at 3–25 mm, and 6.46%, 4.87%, 27.81%, 1.48%, −5.86% at 25–47 mm. It could be found that v_p’s moving velocities along the thickness were improved compared with pure water, but the regularity was not very obvious, which might be related to the complex internal heat conduction and error of thermocouples.

Fig. 6.

(a) Cooling speed peaks along the thickness direction; (b) Movement velocities of cooling speed peaks along the thickness direction. (Online version in color.)

According to Fig. 6, the v_p’s line charts at 25 mm and 47 mm were quite different from that of 3 mm in shape, especially at r=45/60 mm. With v_p moving along the thickness direction, the influence of horizontal heat flux increased, resulting in a reduction in the v_p difference within r=0–45 mm. Regardless of the outermost measurement point (r = 60 mm) which were easily affected by side thermal radiation, the horizontal cooling speed uniformity was determined by δ. The smaller δ was, the more uniform the cooling was.   

$$\delta ={\displaystyle \frac{\sqrt{{\displaystyle \frac{{\sum _{i=1}^n\left(v_i-\overline{v}\right)}^2}{n}}}}{n}}$$(12)

  

$$\overline{v}=\frac{v_1{r}_1^2+v_2\left(r_2^2-r_1^2\right)+v_3\left(r_3^2-r_2^2\right)+v_4\left(r_4^2-r_3^2\right)}{r_4^2}$$(13)

Where, n was the number of temperature measurement points to be calculated in the horizontal direction, and the value was 4 here, r_i=15×i mm (1≤i≤4), v_i was v_p at point r_i-1.

The calculation values of δ were shown in Fig. 7. For a distance of 3 mm from the surface, δ were 2.4–4.1, and the values tended to decrease as v_p moved along the thickness. When pure water was used as coolant, the value reduced from 2.420 at the 3 mm to 0.015 at the 25 mm, and further reduced to 0.014 at the 47 mm. The decrease was obvious near the surface, and then it tended to be gentle as v_p extended inward. Considering the phenomenon that δ caused by the surface jet was weak along the thickness, it could be further inferred that when the thickness of the steel plate was large enough, adjusting the layout of nozzle might not affect core cooling uniformity, Compared with nanofluids, pure water had a relatively lower δ value, and it illustrated good cooling uniformity in the horizontal direction. When nanofluids were used as coolants, nanoparticles covered surface with the evaporation of water, resulting in a strong cooling effect near the jet point, and a weak cooling effect far away from the position. This caused a poor cooling uniformity especially when the concentration was≥1.8 vol%. In this study, single jet was used to obtain the result that adding nanoparticles to the coolant would reduce cooling uniformity, but when multi-jets were used, whether the cooling uniformity would decrease needed further study.

Fig. 7.

Cooling uniformities along the length of experimental plate. (Online version in color.)

4. Conclusion

Using 0.4–2.4 vol% Al₂O₃, TiO₂, SiO₂ + water and pure water as coolants, single jet cooling experiments on the free surface of 50 mm thickness steel plate were completed. Based on the measured temperatures, maximum heat fluxes and cooling speeds were calculated, and the times reaching extreme values were counted. The results showed that for Al₂O₃, TiO₂ and SiO₂ nanofluids at low concentration (0.4 vol% –1.2 vol%), the maximum heat fluxes and cooling speeds were increased compared with pure water. For 1.8 vol% and above nanofluids, the cooling capacity was higher within near the jet point, while it was poor at the position far away from the jet point due to the deposition of nanoparticles hindering the heat transfer. Through comparison, the best cooling efficiency could be obtained by using 1.2% vol Al₂O₃ + water. The cooling speed peaks were 8.14%–19.70%, 2.16%–3.48% and 0.74%–1.44% higher than that of pure water respectively, and moved faster along the thickness direction. However, for single jet, the cooling uniformity in the horizontal direction of experimental plate decreased when nanofluids were used as coolants.

Acknowledgements

We acknowledge the financial support from National Key Research and Development Programs of China (Grant No.: 2017YFB0305100).

References

• 1)   A. H.  Nobari,  V.  Prodanovic and  M.  Militzer: Int. J. Heat Mass Transf., 101 (2016), 1138.
• 2)   A. K.  Mozumder,  M.  Monde,  P. L.  Woodfield and  M. A.  Islam: Int. J. Heat Mass Transf., 49 (2006), 2877.
• 3)   J.  Filipovic,  F. P.  Incropera and  R.  Viskanta: Exp. Heat Transf., 8 (1995), 97.
• 4)   A. K.  Mozumder,  P. L.  Woodfield,  M. A.  Islam and  M.  Monde: Int. J. Heat Mass Transf., 50 (2007), 1559.
• 5)   H.  Fujimoto,  Y.  Shiramasa,  K.  Morisawa,  T.  Hama and  H.  Takuda: ISIJ Int., 55 (2015), 1994.
• 6)   A. M.  Kuraan,  S. I.  Moldovan and  K.  Choo: Int. J. Heat Mass Transf., 108 (2017), 2211.
• 7)   S. J.  Yi,  M.  Kim,  D.  Kim,  H. D.  Kim and  K. C.  Kim: Int. J. Heat Mass Transf., 102 (2016), 691.
• 8)   P.  McInturff,  M.  Suzuki,  P.  Ligrani,  C.  Nakamata and  D. H.  Lee: Int. J. Heat Mass Transf., 127 (2018), 585.
• 9)   Y. L.  He and  Z. X.  Wen: Appl. Therm. Eng., 113 (2017), 1024.
• 10)   J.  Zhang,  X. W.  Zhu,  M. E.  Mondejar and  F.  Haglind: Renew. Sustain. Energy Rev., 101 (2019), 305.
• 11)   S. U. S.  Choi and  J. A.  Eastman: Proc. ASME Int. Mechanical Engineering Cong. and Exposition, ASME FED, New York, (1995), 99.
• 12)   M. H.  Ahmadi,  A.  Mirlohi,  M. A.  Nazari and  R.  Ghasempour: J. Mol. Liq., 265 (2018), 181.
• 13)   S. M. S.  Murshed,  K. C.  Leong and  C.  Yang: Int. J. Therm. Sci., 44 (2005), 367.
• 14)   S.  Aberoumand and  A.  Jafarimoghaddam: J. Taiwan Inst. Chem. Eng., 71 (2017), 315.
• 15)   T.  Yousefi,  E.  Shojaeizadeh,  H. R.  Mirbagheri,  B.  Farahbaksh and  M. Z.  Saghir: Exp. Therm. Fluid Sci., 50 (2013), 114.
• 16)   J.  Lv,  C.  Hu,  M.  Bai,  K.  Zeng,  S.  Chang and  D.  Gao: Exp. Therm. Fluid Sci., 84 (2017), 39.
• 17)   Q.  Li,  Y.  Xuan and  F.  Yu: Appl. Therm. Eng., 36 (2012), 426.
• 18)   C. T.  Nguyen,  N.  Galanis,  G.  Polidori,  S.  Fohanno,  C. V.  Popa and  A.  Le Bechec: Int. J. Therm. Sci., 48 (2009), 401.
• 19)   M.  Chandrasekar,  S.  Suresh and  A.  Chandra Bose: Exp. Therm. Fluid Sci., 34 (2010), 210.
• 20)   H. U.  Kang,  S. H.  Kim and  J. M.  Oh: Exp. Heat Transf., 19 (2006), 181.
• 21)   R.  Prasher,  D.  Song,  J.  Wang and  P. E.  Phelan: Appl. Phys. Lett., 89 (2006), 133108.
• 22)   S. M. S.  Murshed,  K. C.  Leong and  C.  Yang: Int. J. Therm. Sci., 47 (2008), 560.
• 23)   S. E. B.  Maïga,  S. J.  Palm,  C. T.  Nguyen, G. Roy and N. Galanis: Int. J. Heat Fluid Flow, 26 (2005), 530.
• 24)   F. S.  Alkasmoul,  M. T.  Al-Asadi,  T. G.  Myers,  H. M.  Thompson and  M. C. T.  Wilson: Int. J. Heat Mass Transf., 126 (2018), 639.
• 25)   A.  Sergis and  Y.  Hardalupas: Nanoscale Res. Lett., 6 (2011), 391.
• 26)   J. C.  Maxwell: A Treatise on Electricity and Magnetism, Vol. 1, Clarendon Press, London, (1881), 30.
• 27)   P.  Selvakumar and  S.  Suresh: Exp. Therm. Fluid Sci., 40 (2012), 57.
• 28)   K.  Narayan Prabhu and  A. A.  Ashish: Mater. Manuf. Process., 17 (2002), 469.
• 29)   R.  Dou,  Z.  Wen,  G.  Zhou,  X.  Liu and  X.  Feng: Appl. Therm. Eng., 62 (2014), 738.
• 30)   K.  Vivek and  S.  Jahar: Powder Technol., 345 (2019), 717.
• 31)   A. M.  Tiara,  S.  Chakraborty,  I.  Sarkar,  A.  Ashok,  S.  Pal and  S.  Chakraborty: Exp. Therm. Fluid Sci., 89 (2017), 295.
• 32)   M.  Akmal,  A. M. T.  Omar and  M. S.  Hamed: Int. J. Microstruct. Mater. Prop., 3 (2008), 654.
• 33)   T.  Fu,  X.  Deng,  G.  Liu,  Z.  Wang and  G.  Wang: Int. J. Precis. Eng. Manuf., 17 (2016), 1503.