International Journal of Heat and Mass Transfer 85 (2015) 824–829

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Experimental study on pool boiling heat transfer in gradient metal foams

Z.G. Xu, C.Y. Zhao ⇑
Institute of Engineering Thermophysics, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

a r t i c l e i n f o a b s t r a c t

Article history: Pool boiling heat transfer of metal foams with gradient pore densities has been investigated at atmo-
Received 28 October 2014 spheric pressure in saturated deionized water. The gradient foams are made of two foam layers. The foam
Received in revised form 9 February 2015 materials are copper and nickel. Pore densities of foam layers are from 5 PPI to 100 PPI, while the porosity
Accepted 9 February 2015
remains the fixed value of 0.98. A parametric study is performed by varying the foam pore size and
thickness. The effects of alumina nanoparticles and surfactant (SDS) on pool boiling heat transfer of
gradient metal foams are also investigated in the present study. Images of foam fiber deposited by
Keywords:
nanoparticles are captured by SEM. Compared to single-layer foams, gradient metal foams significantly
Pool boiling
Gradient metal foam
enhance pool boiling heat transfer and the enhancement is heavily dependent on foam thickness and
Morphology pore density gradient. The effects of SDS and nanoparticles on pool boiling heat transfer of gradient foams
Nanoparticle are dependent on concentrations.
Surfactant Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction selectively electrodeposited on fin tops of open microchannels.

Recently, Mori et al. [8] found CHF is enhanced by the attachment
Heat flux on integrated circuits has increased to 100 W cm2 in of a honeycomb porous plate to a modified heated surface by
2008 [1], it will increase up to 1000 W cm2 in the near future [2]. nanoparticle deposition.
Heat sink directly affects electronic equipments’ performance and In recent years, open-celled metal foam has been extensively
service life. Therefore, to develop highly efficient heat exchangers studied in phase change heat transfer area due to its high surface
for cooling electronic components is an urgent task. Compared area and other advantages [9]. The experimental results by Xu
with flow boiling, pool boiling needs less additional equipment et al. [10–11] showed that pore density has a more significant
and is an ideal cooling method for equipment system. Compared effect on heat transfer performance than porosity. Zhu et al. [12]
with traditional enhancement surfaces such as machining fins, por- found that the presence of oil reduced nucleate pool boiling heat
ous surfaces with sintered, sprayed or electrodeposited porous transfer on copper foam covers by 15% (maximum) for VG68. The
materials have more surface area and nucleation sites. On the other authors [13–14] of the present study speculated that bubbles
hand, capillary forces of porous materials supply fresh liquid to inside uniform metal foams suffer resistances by metal skeleton
vaporize. Many researchers devoted their effort to investigate boil- and the fresh liquid flowing into vaporization zone during boiling.
ing heat transfer enhancement of porous surfaces. Chien et al. [3] Especially when uniform metal foam is thick or dense, the
found a porous copper particle layer has a higher boiling enhance- resistance slowing down bubble escaping velocity is particularly
ment ratio in methanol or FC-72 than in water. Mori et al. [4] found obvious, resulting in boiling deterioration. Cutting grooves can
critical heat flux (CHF) is enhanced by a honeycomb porous plate improve boiling heat transfer. However it destroys metal foam ori-
because of liquid inflow through the vapor escape channels from ginal structures. Gradient metal foams have high surface area and
the top surface. Zhang et al. [5] demonstrated that alumina can provide suitable space for escaping growing bubbles due to
sponge-like nano-porous structure enhanced pool boiling heat enlarging pores. Li et al. [15–16] experimentally studied critical
transfer. Dai et al.[6] developed a sintered single-layer copper heat flux and heat transfer coefficient of de-ionized water pool
mesh screen to enhance the capillary evaporation heat transfer boiling on gradient porous structures by sintering small particles.
coefficient and CHF. Patil and Kandlikar [7] found that CHF is But the inner structures in gradient metal foams are different from
improved and wall superheat is reduced by microporous coatings gradient stacked particles, and then their boiling heat transfer
mechanisms are not similar. Therefore, the conclusions in Li et al.
[15–16] are not suitable for gradient metal foams. Taking into
⇑ Corresponding author. account the development urgency of heat transfer enhancement
E-mail address: changying.zhao@sjtu.edu.cn (C.Y. Zhao). technology, it is necessary to investigate boiling heat transfer of

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2015.02.017
0017-9310/Ó 2015 Elsevier Ltd. All rights reserved.
 Z.G. Xu, C.Y. Zhao / International Journal of Heat and Mass Transfer 85 (2015) 824–829 825

gradient metal foams. Nanoparticle, surfactant and foam pore size, 3. Results and discussion
thickness effects have been studied in the present study.
3.1. Morphology effect
2. Experiment and metal foam parameters
Pool boiling curves of gradient nickel foams with the pore den-
The whole experimental facility consists of a heating system, a sity variation from 20 PPI to 5 PPI are shown in Fig. 2(a, b). In the
cooling system and a data acquisition system (Fig. 1). The present study, GF-20 PPI–2 mm & 5 PPI–4 mm-Ni represents a gra-
water/nanofluid in the chamber is heated by the two auxiliary hea- dient nickel foam made of 20 PPI down foam layer with the thick-
ters to its saturation point, and then cooled down to the surround- ness of 2 mm and 5 PPI up foam layer with the thickness of 4 mm,
ing temperature naturally to remove the non-condensable gas SF-20 PPI–2 mm–Ni represents a single-layer nickel foam with the
outside. In the experiment, the two auxiliary heaters are first thickness of 2 mm, and similar specifications are used for other
switched on to keep the working water/nanofluid saturated, then cases. It can be found that, sintering 5 PPI nickel foam on 20 PPI
the main heater is switched on to heat the sample substrate from nickel foam surface remarkably enhances the latter’s pool boiling
the beginning heat flux of 8  103 W m2. The thermocouple sig- heat transfer. The up foam layer with a low pore density provides
nals are collected by the Keithley data acquisition unit when their bigger connected pores more space for escaping bubbles from the
variation is within 0.2 K per 20 min. The above procedure is repeat- lower foam layer surface, increases collision time between bubbles
ed when the heat flux increases by the step of 5  104 to and foam skeleton and prolongs bubble growth period inside
10  104 W m2. The maximum uncertainty of is 8.9% by perform- foams. According to the fiber diameter and surface area density
ing the standard uncertainty analysis. More details of the experi- based on the microstructure model of uniform metal foams [18],
ment can be found in the authors’ previous study [17]. In the total surface area of a gradient metal foam can be calculated as:
present study, foam layers were sintered together by Ag–Cu alloy X
n qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sheets in a high-temperature muffle furnace to form gradient met- Stotal  V i 4ðp þ p2 Þð1  ei Þ=dpi ð1Þ
al foams (Fig. 1). Foam porosity is fixed as 0.98, and thickness is i¼1

from 6 mm to 20 mm. The foam material is copper and nickel. where n is foam layer number, e is a foam layer porosity. One foam
Single-layer metal foams are considered as the reference samples layer volume V and pore diameter dp can be respectively calculated
for the comparison purpose. Images of deposited nanoparticle lay- as:
ers on foam fibers were captured by SEM (Sirion 200, Instrument
Analysis Center of SJTU). V ¼ 0:025  0:025  d ð2Þ

Fig. 1. Sketch of the experimental facility and gradient metal foams (all dimensions are in mm).
826 Z.G. Xu, C.Y. Zhao / International Journal of Heat and Mass Transfer 85 (2015) 824–829

dp  0:0254=x ð3Þ 1.2x10

6

where d is a foam layer thickness, x is the foam layer pore density. 6

1.0x10
Then total surface area of a gradient metal foam is:
5
X
n 8.0x10
Stotal  0:025  0:025  di  xi

q(W· m )
-2
i¼1 5
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 6.0x10
 4ðp þ p2 Þð1  ei Þ=0:0254 ð4Þ 5
4.0x10
GF-20 PPI-2 mm & 5 PPI-8 mm-Ni
According to Eq. (4), surface area increments by sintering upper GF-20 PPI-4 mm & 5 PPI-8 mm-Ni
5
foam layers 5 PPI–4 mm, 5 PPI–6 mm and 5 PPI–8 mm on the foam 2.0x10 GF-20 PPI-6 mm & 5 PPI-8 mm-Ni
layer 20 PPI–2 mm are 50%, 75% and 100%, respectively. However, SF-5PPI-8 mm-Ni
smooth copper plate
from the result shown in Fig. 2(a), it can be found that surface area 0.0
0 10 20 30
increment has little effect on the pool boiling heat transfer
ΔTsat(K)
enhancement degree. Surface area increment by sintering upper
foam layers 5 PPI–4 mm, 5 PPI–6 mm and 5 PPI–8 mm on the foam (a) upper layer = 8 mm (20 PPI & 5 PPI)
layer 20 PPI–8 mm are 16.67%, 25% and 33.3%, respectively.
Fig. 2(b) indicates that pool boiling heat transfer enhancement 6
1.2x10
degree firstly decreases and then increases with increasing foam
thickness. The maximum pool boiling heat transfer coefficient 6
1.0x10
increment by the nickel foam 5 PPI–8 mm is about 500% at the heat
flux of 5  105 W m2. Surface area increments by sintering the 5
foam layer 5 PPI–8 mm on the foam layers 20 PPI–2 mm, 20 PPI– 8.0x10
4 mm, 20 PPI–6 mm are 100%, 50% and 33.3%. The results shown q(W· m )
-2
5
in Fig. 3(a) indicate that the pool boiling heat transfer enhance- 6.0x10
ment degree increases with increasing 20 PPI foam thickness. 5
According to Eq. (4), surface area increment by sintering the 4.0x10
GF-100 PPI-2 mm & 5 PPI-8 mm-Ni
upper foam layer 5 PPI–8 mm on the foams layers 100 PPI– 5
GF-100 PPI-4 mm & 5 PPI-8 mm-Ni
2.0x10 GF-100 PPI-6 mm & 5 PPI-8 mm-Ni
SF-5 PPI-8 mm-Ni
smooth copper plate

1.2x10
6 0.0
0 10 20 30
ΔTsat(K)
6
1.0x10
(b) upper layer =8 mm (100 PPI & 5 PPI )
5
8.0x10
Fig. 3. Lower foam layer thickness effect.
q(W· m )
-2

5
6.0x10
2 mm,100 PPI–4 mm,100 PPI–6 mm are 20%, 10%, 6.67%, respec-
5
4.0x10 tively, which indicates surface area increment decreases. Pool boil-
GF-20 PPI-2 mm & 5 PPI-4 mm-Ni
GF-20 PPI-2 mm & 5 PPI-6 mm-Ni
ing heat transfer of GF-100 PPI & 5 PPI increases with increasing
5
2.0x10 GF-20 PPI-2 mm & 5 PPI-8 mm-Ni foam thickness, as shown in Fig. 3(b).Compared with the single
SF-20 PPI-2 mm-Ni
smooth copper plate
foam SF-5 PPI–8 mm-Ni, at low heat fluxes, pool boiling heat trans-
0.0 fer performance of GF-100 PPI & 5 PPI is superior because of higher
0 10 20 30
ΔTsat(K)
surface area. 100 PPI nickel foam has many small pores which
providing a large number of nucleation sites during the boiling
(a) lower layer=2 mm(20PPI&5PPI) process. However, bubbles’ number and volume increases with
increasing heat flux, and they suffer more serious resistance from
1.2x10
6 foam skeleton because of the limited space of connected pores.
Therefore, pool boiling heat transfer coefficient of GF-100 PPI &
1.0x10
6 5 PPI is lower than SF-5 PPI–8 mm-Ni when the heat flux is over
7  105 W m2. Compared the results shown in Fig. 3(b) with those
8.0x10
5 shown in Fig. 3(a), it can be found that, sintering a foam with a low
pore density on a foam with a middle pore density far outstrips a
q(W· m )
-2

6.0x10
5 foam with a high pore density.

5
4.0x10 3.2. Nanoparticle effect
GF-20 PPI-6 mm & 5 PPI-4 mm-Ni
5 GF-20 PPI-6 mm & 5 PPI-6 mm-Ni
2.0x10 GF-20 PPI-6 mm & 5 PPI-8 mm-Ni Fig. 4(a–c) indicate that, adding alumina particle into boiling
SF-20 PPI-6 mm-Ni
smooth copper plate
water worsens pool boiling heat transfer of GF-20 PPI–4 mm &
0.0 5 PPI–8 mm-Cu because the thermal conductivity of alumina
0 10 20 30
(26.8 W m1 K1) is far lower than the copper (398 W m1 K1).
ΔTsat(K)
The 100 nm alumina deposition effect on GF-20 PPI–4 mm &
(b) 5 PPI-8 mm-Cu and GF-100 PPI–4 mm & 5 PPI–8 mm-Cu are shown
lower layer = 6 mm(20 PPI & 5 PPI)
in Fig. 5(a) and (b), respectively. Deposited alumina nanoparticles
Fig. 2. Upper foam layer thickness effect. worsen pool boiling heat transfer of the two gradient copper
 Z.G. Xu, C.Y. Zhao / International Journal of Heat and Mass Transfer 85 (2015) 824–829 827

,
1.4x10
6
X
n
d¼m Si q ð7Þ
6
1.2x10 i¼1

1.0x10
6
where m is the deposited nanoparticle weight, q is the nanoparticle
5 density.
8.0x10
Alumina nanoparticle weight is weight difference between the
q(W· m )
-2

5
6.0x10 copper foams with and without deposited alumina nanoparticles,
5 GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.01%
and alumina nanoparticles on the copper foam fibers can be
4.0x10
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.1% removed by diluted hydrochloric acid. The alumina density is
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.5%
3.98 g cm3. In the present study, the alumina nanoparticle
5
2.0x10 GF-20 PPI-4 mm& 5 PPI-8 mm-Cu

0.0
smooth copper plate
weights on GF-20 PPI–4 mm & 5 PPI–8 mm-Cu and GF-100 PPI–
0 10 20 30 40 4 mm & 5 PPI–8 mm-Cu are 0.08551 g and 0.3868 g, respectively.
ΔTsat(K)
Therefore, based on Eqs. (4) and (7), the mean thicknesses of
(a) 20 nm( 20 PPI=4 mm & 5 PPI=8 mm, Cu) deposited alumina nanoparticle on the two gradient foams are
0.85573 lm and 8.79834 lm, respectively. According to the fiber
6
diameter based on the microstructure model of metal foams [18],
1.4x10
the fiber radius of a foam layer is:
1.2x10
6
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
6
ri ¼ 0:0127  4ð1  ei Þ=ðp þ p2 Þ=xi ð8Þ
1.0x10
5 Based on Eq. (8), the fiber radius of the 100 PPI foam is 63 lm, the
8.0x10
q(W· m )

fiber radius of the 20 PPI foam is 315 lm, the fiber radius of the 5
-2

5
6.0x10 PPI foam is 1260 lm. The length of a foam layer is:
5 GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.01%
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4.0x10
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.1% li ¼ 0:0025  0:025  di  xi  4ðp þ p2 Þ=ð1  ei Þ=ð2pr i  0:0254Þ
5
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.5%
2.0x10 GF-20 PPI-4 mm& 5 PPI-8 mm-Cu
ð9Þ
smooth copper plate
0.0
0 10 20 30 40 Then the average fiber radius of a gradient foam is:
ΔTsat(K)
rmean
(b) 50nm( Pn pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
20 PPI=4 mm & 5 PPI=8 mm, Cu) ¼ 0:0025  0:025  di  xi  4ðp þ p2 Þ=ð1  ei Þ=0:0254
i¼1
¼ Pn pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
i¼1 ¼ 0:0025  0:025  di  xi  4ðp þ p2 Þ=ð1  ei Þ=ð2pr i  0:0254Þ
1.4x10
6 ð10Þ

1.2x10
6 Based on Eq. (10), the average fiber radiuses of GF-20 PPI–4 mm
6
& 5 PPI–8 mm-Cu and GF-100 PPI–4 mm & 5 PPI–8 mm-Cu are
1.0x10
378.1 lm and 195.9 lm, respectively. Therefore, the fiber radius
8.0x10
5 incensement rates of GF-20 PPI–4 mm & 5 PPI–8 mm-Cu and GF-
q(W· m )
-2

5
100 PPI–4 mm & 5 PPI–8 mm-Cu are 0.226% and 4.49%, respective-
6.0x10 ly. Then the surface area incensement rate of GF-20 PPI–4 mm &
4.0x10
5
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.01% 5 PPI–8 mm-Cu is 0.226%, and GF-100 PPI–4 mm & 5 PPI–8 mm-
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.1%
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.5%
Cu is 4.49%. The results in Fig. 5(a) indicate that, in saturated
5
2.0x10 deionized water, pool boiling heat transfer the deposited gradient
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu
smooth copper plate
0.0 foam GF-20 PPI–4 mm & 5 PPI–8 mm-Cu is remarkably lower than
0 10 20 30 40 that in 0.5% boiling alumina nanofluid due to the water’s lower
ΔTsat(K)
effective liquid thermal conductivity. Moreover, deposited alumina
(c) 100nm( 20 PPI=4 mm & 5 PPI=8 mm, Cu) nanoparticles decrease the effective thermal conductivity of cop-
per foam. Furthermore, SEM image shown in Fig. 5(c) reveals that
Fig. 4. Al2O3 nanoparticle concentration effect. deposited nanoparticles form the detached thin layer above the
fiber of gradient foam GF-20 PPI–4 mm & 5 PPI–8 mm-Cu, which
foams. The positive effect of the deposited alumina nanoparticles is significantly increases escaping resistance of the bubbles generat-
the surface area increment. However, alumina thermal conduc- ed on the fibers. Fig. 5(b) reveals that, in saturated deionized water,
tivity is far lower than the copper, which decreases the effective pool boiling heat transfer of the deposited gradient foam GF-
thermal conductivity of the gradient copper foam. Assuming that 100 PPI–4 mm & 5 PPI–8 mm-Cu is only slightly lower than that
metal foam fibers are cylindrical and their average radius is r and in 0.5% boiling alumina nanonfluid. SEM image in Fig. 5(d) shows
the nanoparticle layer thickness is d, the volume increment of a that, the layer sticks to the fiber surface of GF-100 PPI–4 mm &
gradient foam by nanoparticle layers is: 5 PPI–8 mm-Cu and deposited nanoparticles forms a lot of fissures
which can provide more nucleation sites. The positive effect of
X
n X
n
DV ¼ ½pðri þ dÞ  l  pr 2i  l ¼
2 2
½2pr i dl þ pd l ð5Þ fissures partially offsets the negative effects of the lower liquid
i¼1 i¼1 conductivity of water and the downgrading copper foam effective
thermal conductivity by the deposited alumina nanoparticles.
When r i >> d, Eq. (5) can be simplified as:
X
n X
n 3.3. Surfactant effect
DV ¼ 2pr i dl ¼ Si d ¼ m=q ð6Þ
i¼1 i¼1
Fig. 6(a) indicates that, pool boiling heat transfer of GF-40 PPI–
Then the average nanoparticle thickness d on the foam fiber can 6 mm & 20 PPI–8 mm-Ni is heavily enhanced by 100 ppm SDS, and
be calculated as: the maximum boiling heat transfer coefficient increment is more
828 Z.G. Xu, C.Y. Zhao / International Journal of Heat and Mass Transfer 85 (2015) 824–829

6
1.4x10
6
1.2x10
6 1.0x10
6
1.0x10 5
8.0x10
5
8.0x10
q(W· m )
-2

q(W· m )
6.0x10

-2
5
6.0x10
5 5
4.0x10 4.0x10
GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-0.01%
5 GF-40 PPI-6mm & 20 PPI-8 mm-Ni-100 ppm
2.0x10 GF-20 PPI-4 mm& 5 PPI-8 mm-Cu-after deposition
GF-40 PPI-6mm & 20 PPI-8 mm-Ni-400 ppm
GF-20 PPI-4 mm& 5 PPI-8mm-Cu 5
smooth copper plate
2.0x10 GF-40 PPI-6mm & 20 PPI-8 mm-Ni-800 ppm
0.0 GF-40 PPI-6mm & 20 PPI-8 mm-Ni
0 10 20 30 40 smooth copper plate
ΔTsat(K) 0.0
(a) 100nm Al2O3 nanoparticle deposition 0 10 20 30 40
ΔTsat(K)
(δ20 PPI=4 mm & δ5 PPI=8 mm, Cu)

1.4x10
6
(a) 40 PPI=6 mm & 20 PPI=8 mm, Ni
6
1.2x10
6
1.0x10 6
1.0x10 CHF
5
8.0x10
q(W· m )
-2

5
6.0x10
5 8.0x10
5 CHF
4.0x10 5

q(W· m )
6.0x10
GF-100 PPI-4 mm& 5 PPI-8 mm-0.5% -2
5 GF-100 PPI-4 mm& 5 PPI-8 mm-after deposition
2.0x10
GF-100 PPI-4 mm& 5 PPI-8mm
5
0.0
smooth copper plate 4.0x10
0 10 20 30 40 GF-100PPI-6mm & 5PPI-8mm-Ni-100ppm
ΔTsat(K) 5
GF-100PPI-6mm & 5PPI-8mm-Ni-400ppm
2.0x10 GF-100PPI-6mm & 5PPI-8mm-Ni-800ppm
(b) 100nm Al2O3 nanoparticle deposition GF-100PPI-6mm & 5PPI-8mm-Ni
(δ100 PPI=4 mm & δ5 PPI=8 mm, Cu) smooth copper plate
0.0
0 10 20 30 40
ΔTsat(K)

(b) 100 PPI=6 mm & 5 PPI=8 mm, Ni

6
1.2x10
CHF
6
1.0x10

5
8.0x10
CHF
q(W· m )
-2

5
(c) SEM images after boing in 0.5vol% 100nm Al2O3 6.0x10
nanofluid (δ20 PPI=4 mm & δ5 PPI=8 mm, Cu)
5
4.0x10
GF-100PPI-6mm & 20PPI-6mm & 5PPI-8mm-100ppm
5 GF-100PPI-6mm & 20PPI-6mm & 5PPI-8mm-400ppm
2.0x10 GF-100PPI-6mm & 20PPI-6mm & 5PPI-8mm-800ppm
GF-100PPI-6mm & 20PPI-6mm & 5PPI-8mm
smooth copper plate
0.0
0 10 20 30 40
ΔTsat(K)

(c) 100 PPI=6 mm & 20 PPI=6 mm & 5 PPI=8 mm, Ni

Fig. 6. SDS effect.

(d) SEM images after boing in 0.5vol% 100nm Al2O3 20 PPI–8 mm-Ni becomes deteriorated in the 800 ppm SDS solu-
nanofluid (δ100 PPI=4 mm & δ5 PPI=8 mm, Cu) tion. when the heat flux is over 9.2  105 W m2, pool boiling heat
transfer of GF-40 PPI-6 mm & 20 PPI–8 mm-Ni becomes deteriorat-
Fig. 5. Al2O3 nanoparticle deposition effect.
ed in the 400 ppm SDS solution. SDS seriously worsens the pool
boiling heat transfer of GF-100 PPI–6 mm & 5 PPI–8 mm-Ni, as
shown in Fig. 6(b). The connected pores of 100 PPI foam is very
than three times. However, pool boiling heat transfer enhancement small. When bubbles are escaping, the resistance suffering from
of GF-40 PPI–6 mm & 20 PPI–8 mm-Ni decreases with increasing foam skeleton will be very sensitive with the liquid viscosity incre-
SDS concentration because of the increasing liquid viscosity which ment. Based on the equation of Chu et al. [19], the expression for
prevents bubbles escaping. Even worse, when the heat flux is over CHF on a heated horizontal surface with microstructures can be
7  105 W m2, pool boiling heat transfer of GF-40 PPI–6 mm & obtained
 Z.G. Xu, C.Y. Zhao / International Journal of Heat and Mass Transfer 85 (2015) 824–829 829

  1=2
1=4 1 þ cos h 2 p References
qCHF ¼ Chfg q1=2
m ½rlm gðql  qm Þ þ ð1 þ cos hÞ
16 p 4
[1] L. Ping, J. Qian, D. Xu, Overview on electronic equipment thermal control
edql hfg dAW technology, Electro-Mech. Eng. 24 (2008) 1–10.
þ ð11Þ [2] M. Ohadi, Thermal Management of Next Generation Low Volume Complex
Aheat dt
Electronics, www.vita.com/cool/pre/0845-Ohadi.pdf.
The results shown by Wu et al. [20] indicated that equilibrium [3] L. Chien, S. Lee, H. Wang, Effects of fluid properties and surface parameters on
pool boiling of porous and pin-fin surfaces. In: Proceedings of MNHT2008
contact angle is slightly influenced by SDS, while equilibrium Micro/Nanoscale Heat Transfer International Conference January 6–9, 2008,
surface tension decreases with increasing of SDS concentration. Tainan, Taiwan.
Equation (11) indicates that CHF deceases with increasing SDS [4] S. Mori, L. Shen, K. Okuyama, Effect of cell size of honeycomb porous plate
attached to a heated surface on CHF in saturated pool boiling. In: Proceedings
concentration. The experimental results of Fig. 4(b) shows CHFs of the 14th International Heat Transfer Conference, August 8–13, 2010,
of GF-100 PPI–6 mm & 5 PPI–8 mm-Ni in 400 ppm and 800 ppm Washington, DC, USA.
are 9  105 W m2 and 6  105 W m2, respectively, which verifies [5] B.J. Zhang, K.J. Kim, H. Yoon, Enhanced heat transfer performance of
alumina sponge-like nano-porous structures through surface wettability
the above conclusion. Adding 100 ppm SDS into water slightly
control in nucleate pool boiling, Int. J. Heat Mass Transfer 55 (2012) 7487–
influences pool boiling heat transfer of GF-100 PPI–6 mm & 7498.
20 PPI–6 mm & 5 PPI–8 mm-Ni, as shown in Fig. 6(c). However, [6] X. Dai, F. Yang, R. Yang, Y. Lee, C. Li, Micromembrane-enhanced capillary
evaporation, Int. J. Heat Mass Transfer 64 (2013) 1101–1108.
pool boiling heat transfer deterioration degree remarkably increas-
[7] C.M. Patil, S.G. Kandlikar, Pool boiling enhancement through microporous
es with increasing SDS concentration, and CHF reaches earlier, coatings selectively electrodeposited on fin tops of open microchannels, Int. J.
which also verifies the conclusion drawn from Eq. (11). Heat Mass Transfer 79 (2014) 816–828.
[8] S. Mori, S.M. Aznam, K. Okuyama, Enhancement of the critical heat flux in
saturated pool boiling of water by nanoparticle-coating and a honeycomb
4. Conclusions porous plate, Int. J. Heat Mass Transfer 80 (2015) 1–6.
[9] C.Y. Zhao, Review on thermal transport in high porosity cellular metal foams
with open cells, Int. J. Heat Mass Transfer 55 (2012) 3618–3632.
Pool boiling heat transfer of nickel and copper foams with gra- [10] J. Xu, X. Ji, W. Zhang, et al., Pool boiling heat transfer of ultra-light
dient pore densities has been experimentally investigated at atmo- copper foam with open cells, Int. J. Multiphase Flow 34 (2008) 1008–
spheric pressure in saturated deionized water. The main 1022.
[11] Y.P. Yang, X.B. Ji, J.L. Xu, Pool boiling heat transfer on copper foam covers with
conclusions are: Foam thickness and pore density significantly
water as working fluid, Int. J. Therm. Sci. 49 (2010) 1227–1237.
influence pool boiling heat transfer performance of gradient metal [12] Y. Zhu, H.T. Hu, G.L. Ding, H. Peng, X.C. Huang, D.W. Zhuang, J. Yu, Influence of
foams because of bubble escaping resistance variation. Adding alu- oil on nucleate pool boiling heat transfer of refrigerant on metal foam covers,
mina particle into boiling water worsens pool boiling heat transfer Int. J. Refrigeration 34 (2011) 509–517.
[13] Z.G. Xu, Z.G. Qu, C.Y. Zhao, W.Q. Tao, Pool boiling heat transfer on open-celled
of the gradient copper foams because the former’s low thermal metallic foam sintered surface under saturation condition, Int. J. Heat Mass
conductivity. For GF-40 PPI–6 mm & 20 PPI–8 mm-Ni, pool boiling Transfer 54 (2011) 3856–3867.
enhancement degree by SDS decreases with increasing SDS con- [14] Z.G. Xu, Z.G. Qu, C.Y. Zhao, W.Q. Tao, Experimental study of pool boiling heat
transfer on metallic foam surface with U-shaped and V-shaped grooves, J.
centration due to the liquid viscosity preventing bubbles escaping. Enhanced Heat Transfer 19 (2012) 549–559.
For GF-100 PPI–6 mm & 5 PPI–8 mm-Ni and GF-100 PPI–6 mm & [15] C. Li, G. Peterson, Experimental study of enhanced nucleate boiling heat
20 PPI–6 mm & 5 PPI–8 mm-Ni, SDS worsens pool boiling heat transfer on uniform and modulated porous structures, Front. Heat Mass
Transfer 1 (2010) 13654.
transfer. [16] C. Li, T. Li, P. Hodgins, et al., Comparison study of liquid replenishing impacts
on critical heat flux and heat transfer coefficient of nucleate pool boiling on
multi-scale modulated porous structures, Int. J. Heat Mass Transfer 54 (2011)
Conflict of interest
3146–3155.
[17] Z.G. Xu, C.Y. Zhao, Influence of nanoparticles on pool boiling heat transfer in
None declared. porous metal, Appl. Therm. Eng. 65 (2014) 34–41.
[18] Y.P. Du, C.Y. Zhao, Y. Tian, Z.G. Qu, Analytical considerations of flow boiling
heat transfer in metal-foam filled tubes, Heat Mass Transfer 48 (2012) 165–
Acknowledgments 173.
[19] K.H. Chu, R. Enright, E.N. Wang, Structured surfaces for enhanced pool boiling
This work is supported by Shanghai Natural Science Foundation heat transfer, Appl. Phys. Lett. 100 (2012) 214603.
[20] W.T. Wu, Y.M. Yang, J.R. Maa, Nucleate pool boiling enhancement by means of
(15ZR1423400) and the National Key Basic Research Program of surfactant additives, Exp. Therm. Fluid Sci. 18 (1998) 195–209.
China (973 Project: 2013CB228303).