Experimental Thermal and Fluid Science 61 (2015) 259268

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science

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

Experiment study on the burning rates of ethanol square pool res

affected by wall insulation and oblique airow
Pei Zhu a, Xishi Wang a,b,, Changfa Tao a,c
a

State Key Lab. of Fire Science, University of Science and Technology of China, Hefei 230026, China
Anhui Province Center of Collaborative Innovation for City Public Security, Hefei 230026, China
c
Hefei General Machinery Research Institute, Hefei 230031, China
b

a r t i c l e

i n f o

Article history:
Received 25 August 2014
Received in revised form 21 October 2014
Accepted 7 November 2014
Available online 15 November 2014
Keywords:
Pool re
Burning rate
Wall insulation
Oblique airow
Heat transfer

a b s t r a c t
The effects of pool wall insulation condition and oblique air ow on the burning rates of ethanol square
pool res have been studied experimentally in a small-scale wind tunnel, which focused on the heat
transfer process of the fuel pool system due to ame tilt and wall insulation. The results showed that
the mass burning rate of a pool re decreased when the pool wall was insulated, especially in the cases
with larger airow speeds. The maximum value of the mass burning rate without wall insulation reached
approximately 2.3 times of that with wall insulation at an air speed of 2.93 m/s and a tilt angle of 0. For
both cases with or without wall insulation, the mass burning rates decreased with increase in pool sizes
and increased with increase in airow tilt angles. However, for the cases with wall insulation, the differences in the mass burning rates with different pool sizes are relatively small as the airow speeds exceed
1.5 m/s, while the differences are relatively large for the cases without wall insulation. In addition, the
turbulence of the ame uid would be weakened when the pool wall was insulated.
2014 Elsevier Inc. All rights reserved.

1. Introduction
The re hazards of liquid hydrocarbon fuels have been a serious
safety concern, which is attributed to liquid hydrocarbons being
ubiquitous throughout society, including in oil tanks, etc. A considerable number of studies [17] on both the practical and the fundamental aspects of pool res have been conducted, which
focused on the burning rate, the ame height and pulsation, the
thermal radiation, the total kinetic energy, and the soot microstructure. However, the above studies mainly considered the cases
of burning in a quiescent ambient condition. In practical scenes,
most pool res may burn under airow environments, such as
ambient wind in storage plants, etc.
Some reports had been focused on the effects of airow on
pool res [812]. For a large pool re, the inuence of wind
speed on the burning rate was negligible up to 2 m/s, but it
had a certain effect at wind speeds higher than 2 m/s [8]. The
results [9] of methanol square pool res with transverse air
speeds from 0 to 5.5 m/s showed that the burning rate of the
smallest square pool (7.5  7.5 cm) monotonously increased with
Corresponding author at: State Key Lab. of Fire Science, University of Science
and Technology of China, Hefei 230026, China. Tel.: +86 551 63606437; fax: +86
551 63601669.
E-mail address: wxs@ustc.edu.cn (X. Wang).
http://dx.doi.org/10.1016/j.expthermusci.2014.11.006
0894-1777/ 2014 Elsevier Inc. All rights reserved.

increased wind speeds, while the burning rate of the largest

square pool (30  30 cm) was essentially invariant to this range
of air speeds, and the intermediate-sized square pools showed
a non-monotonic increase. The experiment of a small pool re
in a 1/20 reduced-scale tunnel [1012] showed that the burning
rates of methanol square pool res decreased, while those of
acetone and heptane square pool res increased with increased
air speeds.
It should be noted that in most previous studies, the burning
pools were embed into the oor, so that the pool wall was not
exposed to the surroundings and the tilted ame could not touch
the pool wall. But in most practical scenes, the burning pools
may be placed above the oor, such as the oil tanks in petrochemical enterprises, etc. In these scenarios, the tilted ame can touch
the pool wall and this will enhance the heat transfer from pool
to the fuel and then accelerate the fuel evaporation.
However, most of the above studies focused on horizontal cross
airow, but few considered the effect of tilt wind on re behaviour.
Several studies [13,14] had been investigated the effect of tilt wind
on re behaviours, but only focused on the ame spread rate, ame
shape of solid fuel res. The studies [15,16] on the effects of oblique air ow on burning rates of square ethanol pool res showed
that the mass burning rate increased with the increase of tilt direction and the pool wall heat effect became more obvious under oblique airow, but they did not consider the effect of the pool wall

260

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

Nomenclature
L
A
cp
m
_0
m
_ 00
m
_ 00
Dm

pool edge length (cm)

fuel surface area (m2)
specic heat (kJ kg1 K1)
fuel mass (g)
mass burning rate (gs1)
mass burning rate per unit area (gm2 s1)
difference between with and without wall insulation
(gm2 s1)
_ 00tilted
mass burning rate per unit area with oblique airow
m
(gm2 s1)
_ 00lev el
mass burning rate per unit area with horizontal airow
m
(gm2 s1)
_ 00tilted =m
_ 00lev el
k
value of m
h
tilt angle ()
Hg
gasication heat of the fuel (kJ kg1)
Hv
latent heat of evaporation (kJ kg1)
net energy to evaporate the fuel (kW)
Q_ ev p
Q_ ev pwithout net energy to evaporate the fuel without wall insulation (kW)
Q_ ev pwith net energy to evaporate the fuel with wall insulation
(kW)
total heat transfer from the ame and surroundings to
Q_ in
the fuel (kW)
energy loss from the fuel to the surroundings (kW)
Q_ out
radiative heat transfer from ame to the fuel surface
Q_ frad
(kW)
convective heat transfer from the ame to the fuel surQ_ fconv
face (kW)

insulation, although it indicated that the insulation may mainly

affect heat transfer [17,18].
The current experimental study focuses on examining if the
wall insulation conditions would inuence the burning characteristics of the pool re under oblique airow due to ame tilt and the
extent of these effects compared to that without wall insulation.
Then attempt to provide method on how to reduce the hazards
of a liquid pool re under such conditions. However, there are
few relevant reports on this topic. The results may be helpful for
estimating the mass loss and heat transfer of a pool re under
the coupling effects of oblique airow and pool wall insulation.

2. Experimental setup and methodology

2.1. Experimental facility and condition
Fig. 1 shows the schematic diagram of the experimental setup.
The cross-section area is 42  60 cm. The fuel used in the experiments was ethanol-C2H5OH, its properties are shown in Table 1.
An ethanol ame generally produces little soot when combusting
under normal atmospheric conditions; therefore, the ame emits
relatively low levels of thermal radiation compared to yellow
soot-producing ames. The stainless steel square pools with edge
lengths, L, of 4, 6, 8, and 10 cm, a wall thickness of 1 mm and a
depth of 1.5 cm were used based on the size in comparison with
the cross-sectional area of the model.
As shown in Fig. 1, airow was produced using a 0.75 kW wind
fan positioned at one end of the wind tunnel, which generated
downward wind with air speed up to 5 m/s. The angle h represents
the wind tilt direction. A KA12 four-channel anemometer with resolution of 0.01 m/s was used to measure the airow speed inside
the wind tunnel. In this work, airow speeds ranging from 0 to

Q_ wfuel
Q_ f w
Q_ rad
Q_ conv
Q_ cond
Q_ loss1
Q_ loss2
v
T
T0
t

total heat transfer from the pool wall to the fuel (kW)
heat transfer from the ame to the pool wall through
the top surface (kW)
radiative heat transfer from the tilt ame to the pool
wall outside surface (kW)
convective heat transfer from the tilt ame to the pool
wall outside surface (kW)
conductive heat transfer from the tilt ame to the pool
wall outside surface (kW)
heat loss from the fuel to the surroundings through the
pool wall (kW)
heat loss from the fuel surface to the surroundings
through reradiation (kW)
airow velocity (m/s)
wall temperature with wall insulation(C)
wall temperature without wall insulation (C)
time (s)

Subscripts
with
with wall insulation
without without wall insulation
f
ame
w
pool wall
l
leeward rim
s
streamside rim
w
windward rim

3.0 m/s were considered to provide relatively steady airow. Turbulence intensity of the airow was found to be less than 6% at
all test points.
The burning pool was placed on a horizontal support frame. An
asbestos board with a thickness of 1 cm between the pool and the
support frame was used to avoid the pool bottom heated by the
ame and reduce heat loss from pool bottom. An electric balance
with sampling intervals of 0.1 s and resolution of 0.01 g was positioned below the support frame to record the mass loss history of
the fuel. Images of the ames were recorded with a 3.3 Mega pixels
SONY HDR-PJ10 digital camera.
The experiments were carried out in a laboratory under the following ambient conditions: the temperature was 18 2 C, the
pressure was 101 5 kPa, and the humidity was 25 15%. All of
the test cases are specied in Table 2. Each test case was carried
out at least three times. In each test, the door and the window were
closed to avoid the inuence of ambient wind, and the fan was
switched on after all of the equipment and the fuel were set to
the prescribed conditions. The fan and the data acquisition system
were activated at the same time, then after approximately 30 s, the
fuel was ignited. Each repeated test was conducted after the wind
tunnel and fuel pan returned to the initial ambient conditions.

2.2. Heat balance of the fuel without pool wall insulation

Fig. 2 shows the schematic diagram of the heat balance in a
liquid fuel pool re without wall insulation under oblique airow.
Three K-type thermocouples with bead diameters of 1 mm and a
resolution of 0.1 C were soldered at the middle of the wall surfaces to measure the wall temperatures, which are according to
the leeward wall temperature, T 0l , stream side wall temperature,
T 0s , and windward wall temperature, T 0w , respectively. The fuel

261

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

Fig. 1. Schematic diagrams of the experimental apparatus.

Table 1
The properties of the fuel.
Fuel

Density

Boiling
point

Heat of
combustion

Specic heat at
constant pressure

Ethanol

0.79 g/cm3

78.4 C

1365.5 kJ/mol

2.4 kJ/(kg C)

evaporating rate is determined by the net gain energy, Q_ ev p , of the

fuel from ame and surroundings, which is the difference between
the total heat Q_ in into the fuel and the total heat loss Q_ out from the
fuel. The heat loss through the bottom of the pool was ignored due
to the asbestos board. Thus, the heat balance in the fuel can be
represented as:

Table 2
Experimental test cases.
Test cases

Initial fuel mass (g)

Tilt angle of airow (h, )

Airow speed v (m/s)

Case 1

17

0
10
20
30

0/0.28/0.49/0.72/1.02/1.25/1.74/2.31/2.93

With and without

Case 2

35

0
10
20
30

0/0.28/0.49/0.72/1.02/1.25/1.74/2.31/2.93

With and without

Case 3

70

0
10
20
30

0/0.28/0.49/0.72/1.02/1.25/1.74/2.31/2.93

With and without

Case 4

95

0
10
20
30

0/0.28/0.49/0.72/1.02/1.25/1.74/2.31/2.93

10

With and without

Fig. 2. Schematic of the pool structure and layout without wall insulation.

L (cm)

Pool wall insulation

262

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

Q_ in Q_ frad Q_ fconv Q_ wfuel

Q_ wfuel Q_ f w Q_ rad Q_ cond Q_ conv

Q_ out Q_ loss1 Q_ loss2

where Q_ frad is radiative heat transfer from ame to fuel, Q_ fconv is

convective heat transfer from ame to fuel, Q_ wfuel is the total heat
transfer from pool wall to fuel, Q_ f w is the conductive heat transfer
from ame to wall top surface, Q_ loss1 is the heat loss from fuel to
the surroundings, Q_ loss2 is the heat loss through pool wall to the surroundings. Q_ rad ; Q_ cond ; Q_ conv are radiation, conductive, convective
heat transfer from ame to the pool walls, which include two
stream side walls and the leeward wall due to ame tilt. Therefore,
the net gain energy Q_ ev p can be expressed as:

Q_ ev pwithout Q_ in  Q_ out
Q_ frad Q_ fconv Q_ f w Q_ rad Q_ cond Q_ conv
4

Hg Hv

cp dT

Density (g/cm3)

Thermal conductivity (w/m k)

Stainless steel
Aerogel Blanket

7.79
0.22

16 (20 C)
0.018 (25 C)

23 (600 C)
0.045 (650 C)

Similarly, the three wall temperatures were also measured, which

are Tl, Ts and Tw respectively. Due to the thermal insulation layer,
the heat transfer of Q_ rad ; Q_ conv ; Q_ cond and Q_ loss1 will be ignored.
Therefore, the heat balance for the fuel can be represented as:

Q_ in Q_ frad Q_ fconv Q_ wfuel

Q_ wfuel Q_ f w

Q_ out Q_ loss2

Q_ ev pwith Q_ in  Q_ out Q_ frad Q_ fconv Q_ f w  Q_ loss2

_ 00with  A  Hg
m

Hg is the effective heat of gasication, which is dened as:

Material

Thus, the net gain energy Q_ ev p can be represented as:

 Q_ loss1  Q_ loss2
_ 00without  A  Hg
m

Table 3
The properties of the pool and thermal insulation material.

R
where Hv is the fuel latent evaporation heat and cpdT is the sensible heat of fuel warming up.
In Eq. (4), the radiative heat transfer Q_ frad from ame to fuel due
to the ame tilt, should be very small, which has been discussed by
Hu et al. [20]. Q_ f w ; Q_ loss1 ; Q_ loss2 are also relative small comparing to
other terms. The forced airow may induce surface circulation of
the liquid, and thus improve the convective heat transfer Q_ fconv at
the gasliquid interface. In addition, the ame wrapping will lead
to the pool walls heating due to Q_ rad ; Q_ cond ; Q_ conv , where the most
of the heat absorbed by the pool walls being considered transfer
to the liquid fuel through convective heat transfer at the solid
liquid interface eventually. So all of these convective heat transfer
will be the main factors that affect the heat ux rate received by
the fuel in the cases without wall insulation.

Due to the wall insulation, the heat of Q_ rad ; Q_ cond ; Q_ conv transferred
from ame to pool wall were weakened obviously, while the convective heat transfer Q_ fconv was still important due to the forced
airow.
Therefore, the difference in the net gain energy of the fuel DQ_ ev p
between the cases without and with wall insulation can be represented as:

DQ_ ev p Q_ ev pwithout  Q_ ev pwith Q_ rad Q_ cond Q_ conv  Q_ loss2

_ 00  A  Hg
Dm

10

_ 00 m
_ 00without  m
_ 00with , and m
_ 00without is the mass burning rate
where Dm
_ 00with is the mass burning rate with wall
without wall insulation and m
insulation. Comparing to the heat of Q_ rad ; Q_ cond , and Q_ conv transferred from the ame to the pool wall, the heat loss q_ loss2 is small.
3. Experimental results and discussion

2.3. Heat balance of fuel with pool wall insulation

3.1. Flame behaviour of a pool re

Fig. 3 shows the schematic diagram of the heat balance in a

liquid fuel pool re with wall insulation under oblique airow.
The pool wall insulation was realized by wrapping a layer of thermal insulation material with 5-mm thickness outside the pool. The
thermal insulation material being used here is Aerogel Blanket,
which is non-ammable, and its properties are shown in Table 3.

Figs. 4 and 5 show the typical ame images of a 10  10 cm pool

re with an increase in airow speeds and tilt angles without and
with wall insulation, respectively. In the quiescent condition, the
ame shapes generally depend on the relative strength of the
buoyancy of the combustion products in drawing ambient air horizontally over the pools base to participate in the combustion. As

Fig. 3. Schematic of the pool structure and layout with wall insulation.

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

263

Fig. 4. Flame images of pool res under different airow speeds and directions (L = 10 cm, without pool wall insulation).

Fig. 5. Flame images of pool res under different airow speeds and directions (L = 10 cm, with pool wall insulation).

the airow speeds increased, even by a small amount, under the

quiescent conditions (e.g., 0.49 m/s), the ame shapes were distorted considerably, with the main part of the ame shifting to
the downstream side, which is because the buoyant force has not
persisted in the ow for a long time, and the horizontal air momentum dominates in the near eld of the pool. In our experiments, the
fuel pan was only put on a support frame whose top was a
10  10 cm at plate, and the leeward edge of the fuel pan was
aligned with the downstream edge of the at plate.
It is found that with a tilt angle of 20 or 30, the ame was
tilted horizontally and even downward, where the tip of the tilted
ame was lower than the bottom of the pool at high airow
speeds. This is a particular characteristic of the ame with the
effect of oblique airow as shown in Figs. 4 and 5. It is also found
that the turbulence intensity and the volume of the ames with
wall insulation weakened and reduced compared to those without
wall insulation. For the cases with pool wall insulation, the wrapping of the ame around the fuel pan side rim was not obvious
when the tilt angle was 0 and 10, but obvious for other cases.
The detachment of the ames from the windward rim of the
pools occurred at relatively large air speeds and tilt angles, which
was obvious at all tilt angles when the airow speed exceeded

1.74 m/s without wall insulation. In addition, the ame length

was also detected based on the ame images, which showed that
the ame length rst increases and then decreases with increased
airow speeds. From these phenomenon, it can be preliminary
deduced that the fuel evaporation and fuel vapour diffusion rate
from the fuel surface decreased with wall insulation. Other explanations will be combined with the measured mass loss and temperature data in the following sections.
3.2. Typical temporal evolution of fuel mass and pool wall temperature
A typical temporal evolution of fuel mass loss m and mass loss
_ 0 with and without wall insulation are shown in Fig. 6, which
rate m
corresponds to a pool size with L = 10 cm, a tilt angle of 30, and an
airow speed of 2.31 m/s. It can be seen that the total burning time
and steady burning time with wall insulation are all longer than
those without wall insulation. Generally, as shown in Fig. 6, the
variation of the burning rate can be divided into four stages. In
the cases with pool wall insulation, they are the rst stage with
rapid increase, the second stage with slow increase, the third stage
with a long time of steady combustion, and the last stage with
rapid decaying. However, in the cases without pool wall insulation,

264

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

3.3. Burning rate of a pool re without and with wall insulation

0.45
100

0.35

mass (g)

80
steady stage

60

0.30
0.25
0.20

40

0.15
20

0.10

mass burning rate (g/s)

0.40

0.05

0
0

100

200

300

400

500

600

0.00

t (s)
Fig. 6. Typical temporal evolution of the mass loss and mass burning rate of pool
res with and without wall insulation (h = 30, v = 2.31 m/s, L = 10 cm).

rapid increase occurred for the second stage, and the steady stage
remained for a relatively short time. These can be explained that as
the heating effect of the three rims of the pool by ame would
decrease remarkably due to wall insulation. However, in the cases
without wall insulation, the pool wall will be heated directly by the
wrapping ame, which will enhance the fuel evaporation rate.
Based on the requirement of data analysis, only the steady burning
stage was in a quasi-steady state system. During this stage, the fuel
liquid level will decrease with ongoing burning, but the temperature readings were still steady at this stage in the current study
for both of the cases.
Fig. 7 shows the typical temporal evolution of the wall temperature, which corresponds to the cases in Fig. 6. In the steady burning stage, the wall temperature for both cases basically remains a
constant value. This will further account for the stability of the
mass burning rate at this stage.
In addition, it can be found that the windward rim temperature
is lower than the other rim temperature because the ame will be
detached from the windward rim at an airow speed of approximately 2.31 m/s (as shown in Figs. 4 and 5), while the temperature
of the side and leeward rims are relative large due to the heating of
the wrapping ame. In the last decay stage, the rapidly rising rim
temperature is caused by the fuel burning out, and the fuel pan
inside can be heated directly by the ame. Therefore, in the following sections, the time-averaged mass burning rates per unit area
_ 00 at the steady stage were selected as the nal mass burning rate
m
in the analysis.

200

With wall insulation Without wall insulation

Tl
Ts

180

T'l
T's

Tw

160

T'w

T (C)

140

steady stage

120
100
80
60
40
20
0

100

200

300

400

500

600

t (s)
Fig. 7. Typical temporal evolution of the pool wall temperature with and without
wall insulation (h = 30, v = 2.31 m/s, L = 10 cm).

_ 00without of a pool re without wall insulation

The burning rates m
under different airow speeds and directions are presented in
Fig. 8. It is obviously shown that the mass burning rates increase
with increased air speeds and decrease with increased pool size,
_ 00 / 1=L. The results related to the horand it can be expressed as m
izontal airow agree well with the results of a small square methanol pool re presented by Wood et al. [9]. For a small tilt angle, as
shown in Fig. 8(a) and (b), the burning rates increase monotonically, and rapid increase occurs approximately 1.5 m/s. However,
for relative larger tilt angles, as shown in Fig. 8(c) and (d), the burning rates vary non-monotonically with the increase of airow
speeds, which involves short-lived stability. When the tilt angle
is relative large, the side and leeward rim of the pool will be heated
by the tilt ame, which cause initial sharp increases in the burning
rate. The short-lived stability may be caused by steady net energy
gain Q_ ev p from the ame and the surroundings of the fuel.
It is also found that as air speeds increase from 1.25 m/s to
1.74 m/s, the mass burning rates increases rapidly at different
wind directions and pool size. As shown in Fig. 4, the ame detachment occurred at the air speed of 1.74 m/s, but it did not happen at
the air speed of 1.25 m/s. When the ame was detached, the turbulence of the ame was obviously enhanced, and the convective
heat transfer between the fuel and the ame at the gasliquid
interface would be improved, therefore the combustion process
would be accelerated. When the air speed exceeds 1.74 m/s, the
increase in the mass burning rates is reduced. Overall, the various
paths for heat transfer from the ame to the fuel lead to the complex coupling effects on fuel evaporation and fuel combustion. The
key point in steady state is that the net rate of energy to the fuel by
all of these mechanisms balances the rate of energy consumed for
fuel evaporation.
_ 00with of pool res with wall insulation under
The burning rates m
different airow speeds and directions are presented in Fig. 9. Similar to Fig. 8, the burning rates decrease with the increase of pool
size, but the burning rates decrease obviously compared to the
cases without wall insulation. In addition, the differences in the
burning rates with different pool sizes are relatively small compared to the cases without wall insulation. In which, the differences are relatively large when the airow speeds exceed 1.5 m/
s. For a small tilt angle (h = 0 and h = 10), as shown in Fig. 9(a)
_ 00with rst decrease and then increase
and (b), the burning rates m
with increased air speeds, except for the cases of 4  4 cm pool
res.
For small tilt angles and low air speeds, ame tilting would
cause the view factor from the ame to the fuel surface to decrease,
which would reduce the radiative heat transfer Q_ frad . In addition,
due to the small vertical components of the airow speeds, the
convective heat transfer Q_ fconv from the ame to the fuel is relative
small. On the other hand, due to wall insulation, the heating effect
by ame wrapping is approximately eliminated. Meanwhile, the
ame is vertical in the quiescent condition, and the energy from
the ame to the fuel will be relatively large compared to the cases
at low air speeds. All of these lead to the rst decrease in the burning rates. With the continued increase in air speeds, the convective
heat transfer Q_ fconv will obviously increase, which will lead the further increases in the burning rate.
When the tilt angle further increases up to 20 and 30, the
_ 00with will always increase with increased airow
burning rates m
speeds, except for the cases where the airow speeds are below
0.28 m/s. The understanding of this phenomenon is that the effects
of the oblique airow on the ame shape and the fuel become
more obvious due to the increase in the vertical components of
the airow speeds. For instance, the convective heat transfer
Q_ fconv was obviously enhanced compared to the one with small tilt

265

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

-2 -1

40

50

=0
4cm
6cm
8cm
10cm

m''(g/m s )

-2 -1

m''(g/m s )

50

30
20

(a)

10
0.0

0.5

1.0

1.5

2.0

2.5

40

=10
4cm
6cm
8cm
10cm

30
20

(b)

10

3.0

0.0

0.5

1.0

v (m/s)

=20

50

4cm
6cm
8cm
10cm

-2 -1

40

m''(g/m s )

-2 -1

m''(g/m s )

50

30
20

(c)

10
0.0

0.5

1.0

1.5

1.5

2.0

2.5

3.0

v (m/s)

2.0

2.5

40

=30
4cm
6cm
8cm
10cm

30
20

(d)

10

3.0

0.0

0.5

1.0

v (m/s)

1.5

2.0

2.5

3.0

v (m/s)

Fig. 8. Burning rate of square pool res without wall insulation for different airow speeds and pool sizes.

=0

28

4cm
6cm
8cm
10cm

24
20
16
12

0.5

4cm
6cm
8cm
10cm

1.0

1.5

2.0

2.5

20
16
12

(a)
0.0

=10

24

m (gs-1m-2)

m ( gs-1m-2)

28

3.0

(b)
0.0

0.5

1.0

-1

m (gs m

=20
4cm
6cm
8cm
10cm

28

m (gs-1m-2)

24

-2

28

20
16

(c)

12
0.0

0.5

1.0

1.5

1.5

2.0

2.5

3.0

v (m/s)

v (m/s)

2.0

2.5

3.0

v (m/s)

24

=30
L=4cm
L=6cm
L=8cm
L=10cm

20
16

(d)

12
0.0

0.5

1.0

1.5

2.0

2.5

3.0

v(m/s)

Fig. 9. Burning rate of square pool res with wall insulation for different airow speeds and pool sizes.

angles of airow. For the cases with larger pool sizes, especially the
8  8 cm and 10  10 cm pools, the lowest value of the burning
rate occurred when the airow speed was approximately 0.28 m/
s. The reasons are similar to the cases with small tilt angles, which
was discussed above.
However, for a small pool size of 4  4 cm with all of the tilt
angles, the mass burning rate increased with increased airow

speeds. It can be interpreted that the heat feedback per unit area
from the ame to the fuel is relatively large compared to the large
pool size. For all of the pool sizes, there is also a critical value of air
speed leading to a relatively fast increase in the burning rate,
which corresponds to the ame detachment. In addition, from
Fig. 9, for all of the pool sizes and oblique airow angles, a steady
or very slow state of increase in the mass burning rate with

266

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

increased airow speeds generally appears, which can also be

interpreted as the net rate of energy gain to the fuel by all of these
mechanisms and balances the rate of energy consumed for fuel
evaporation.
_ 00without =m
_ 00with is introduced to
As shown in Table 4, a parameter m
quantitatively indicate the difference in the mass burning rate
between the cases with and without wall insulation. The values
correspond to the ratio of the mass burning rates with different airow tilt angles and pool sizes at the maximum airow speed of
_ 00without =m
_ 00with is approximately
2.93 m/s. The maximum value of m
2.3 for the 4  4 cm pool with a tilt angle of 0, while the minimum
value is approximately 1.43 for the 10  10 cm pool with a tilt
angle of 30. Therefore, it can be said that the effects of wall insulation on the ame burning rate are very notable for the cases with
relatively large airow speeds and small pool sizes.

3.4. Effects of airow directions on mass burning rates

To indicate the inuence of airow directions on the burning
rate of pool re more clearly, the burning rates with and without
wall insulation for different airow tilt angles were plotted in
Figs. 10 and 11. The data of burning rates for different pool sizes
under different tilt angles and airow speeds are shown in
Fig. 10. It can be seen that the mass burning rates increase with
increased tilt angles for all of the cases. However, the differences

Table 4
The ratio of the mass burning rate per unit area with different pool sizes and airow
tilt angles.

40
30

With wall insulation

=0
=10
=20
=30
Without wall insulation
'=0
'=10
'=20
'=30

_ 00tilted =m
_ 00lev el
km

50

(a)

10
0.0

0.5

1.0

1.5

2.0

2.5

_ 00lev el
m

where
is the mass burning rate with oblique airow,
is
the mass burning rate with horizontal airow.
Figs. 11 and 12 show the results of k with and without wall
insulation for different airow speeds and pool sizes, where kwith
and kwithout represent the data of the pool re with or without wall
insulation, respectively. It can be seen that there are signicant differences in the effect of airow directions on the mass burning rate
between the two cases. In addition, the values of k increase with an
increase in the tilt angle for both cases.
For the cases without wall insulation, as shown in Fig. 11, the
values of k rst increase until a peak value is reached and then
decrease with an increase in airow speed. Combined with
Fig. 10, it can be concluded that the effect of the airow direction
on the mass burning rate is only prominent at relatively lower airow speeds (e.g., 0.491.25 m/s). As shown in Fig. 4, at this range
of airow speeds, the variation of ame shape is obvious with
the increased airow tilt angle. For instance, with an air speed of
0.72 m/s, the ame tilt gradually increased with the airow tilt
angle increasing from 0 to 30. At h = 0, the ames deected
but did not reach the horizontal, and the ame did not completely
wrap the pool wall. However, at h = 30, the ame deected to the
horizontal and would touch the pool wall, such that the energy of
Q_ rad ; Q_ cond , and Q_ conv from the ame to the pool wall would
enhance, thus leading to an increase in the net energy gain Q_ ev p .
Under large airow speeds, the ame would be totally
deected to the horizontal or even downward, the side and

L=4cm

20

11

_ 00tilted
m

m'' (g/m -2 s -1)

m'' (g/m -2 s-1 )

50

in the mass burning rate with different tilt angles are relatively
large in the cases with wall insulation when airow speed exceeds
1.5 m/s. In addition, the difference of burning rate between the
cases without and with wall insulation would decrease with the
increase of pool size. To describe the variation of HRR with ventilation velocity, Carvel et al. [19] dened a coefcient, k = Qvent/Qnat,
where Qvent is the HRR with forced ventilation and Qnat is the HRR
with natural ventilation. Similar to this, we dene the mass burning rate enhancement coefcient, k, as:

40
30

With wall insulation

=0
=10
=20
=30
Without wallinsulation
'=0
'=10
'=20
'=30

20

(b)

10

3.0

0.0

0.5

1.0

v (m/s)

40
30

With wall insulation

=0
=10
=20
=30
Without wall insulation
'=0
'=10
'=20
'=30

1.5

2.0

2.5

3.0

2.5

3.0

v (m/s)
L=8cm

50

m'' (gm2s -1 )

m'' (gs -1 m-2 )

50

L=6cm

20

40
30

With wall insulation

=0
=10
=20
=30
Without wall insulation
'=0
'=10
'=20
'=30

L=10cm

20

(c)

10
0.0

0.5

1.0

1.5

v (m/s)

2.0

2.5

3.0

(d)

10
0.0

0.5

1.0

1.5

2.0

v (m/s)

Fig. 10. Mass burning rate with and without wall insulation for different airow speeds and directions.

267

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

1.5

Without wall insulation

=10

=20

6cm

Without wall insulation

4cm
=30

=10

1.6

=20

=30

1.4
1.4

k without

kwithout

1.3
1.2
1.1

1.2
1.0

1.0

(a)

0.9
0.0

0.5

1.0

1.5

2.0

2.5

(b)

0.8
0.0

3.0

0.5

1.0

v (m/s)

1.5

1.8

2.5

3.0

Without wall insulation 10cm

Without wall insulation 8cm

=10

1.6

2.0

v (m/s)

=20

1.8

=30

=10

=20

=30

kwithout

k without

1.6
1.4
1.2
1.0

1.4
1.2
1.0

(c)

0.8
0.0

0.5

1.0

1.5

2.0

2.5

(d)

0.8
0.0

3.0

0.5

1.0

v (m/s)

1.5

2.0

2.5

3.0

v (m/s)

Fig. 11. A plot of k via airow speed without wall insulation for different airow tilt angles and pool sizes.

1.5

With wall insulation 4cm

=10

=20

1.6

=30

With wall insulation 6cm

=10
=20

=30

1.4
1.4

k with

k with

1.3
1.2

1.2

1.1
1.0
1.0

(a)
0.8
0.0

0.9
0.0

0.5

1.0

1.5

2.0

2.5

3.0

(b)
0.5

1.0

v (m/s)
1.8

1.5

2.0

2.5

3.0

v (m/s)

With wall insulation 8cm

=10

=20

With wall insulation 10cm

=30

1.8

=10

=20

=30

1.6

k with

k with

1.6
1.4
1.2

1.2

1.0
0.8
0.0

1.4

(c)
0.5

1.0

1.5

2.0

2.5

3.0

v (m/s)

1.0
0.8
0.0

(d)
0.5

1.0

1.5

2.0

2.5

3.0

v (m/s)

Fig. 12. A plot of k via airow speed with wall insulation for different airow tilt angles and pool sizes.

leeward rims of the fuel pan would all be wrapped by the ame,
and the differences in the energy from the ame to the wall will
be not large with different airow tilt angles at the same airow
speed. Therefore, the differences in the burning rate with

different airow directions are mainly caused by the convective

heat transfer Q_ fconv from the ame to the fuel. However, the
energy in this part is relatively small compared to wall heated
effect on fuel heating. Therefore, the effect of the airow

268

P. Zhu et al. / Experimental Thermal and Fluid Science 61 (2015) 259268

directions on the mass burning rate is not obvious for the cases
with large airow speeds.
For the cases with wall insulation, as shown in Fig. 12, the values of k rst increase and then remain steady at a certain value
with an increase in airow speeds. Due to the insulation of the pool
wall, the effect of the wall heating will not exist. The effects of
the oblique airow on the convective heat transfer Q_ fconv from
the ame to the fuel would be the main part that would lead to
the change in the burning rates with different airow tilt angles.
Therefore, the differences in the burning rates with different tilt
angles are relatively large with wall insulation, compared to that
without wall insulation under large airow speeds.
4. Conclusions
The effects of oblique airow on the burning rate of square ethanol pool res with and without wall insulation have been studied
experimentally based on a small-scale wind tunnel. The variation
in the fuel mass loss and wall temperature were measured to analyse the effect of wall insulation and oblique airow on the pool
re burning rates. The following conclusions can be drawn:
(1) The mass burning rates are obviously reduced due to the
insulation of the pool wall, especially in the cases with larger
airow speeds.
(2) The mass burning rates decrease with increased pool size.
However, in the cases with wall insulation, the differences
in the mass burning rates are relatively small, while for
the cases without wall insulation, the differences are relatively large, especially when the airow speeds exceed
1.5 m/s.
(3) The burning rates increase with increased airow tilt angles
for both cases with or without wall insulation. However, in
the cases with wall insulation, the change in the mass burning rates with different airow tilt angles are relatively large
when the airow speed exceeds 1.5 m/s.

Acknowledgments
The authors appreciate the support of the China National Key
Basic Research Special Funds Project (Grant No. 2012CB719704)
and the Natural Science Foundation of China (Grant No. 51323010).
References
[1] A. Nasr, S. Suard, H. El-Rabii, J.P. Garo, L. Gay, L. Rigollet, Heat feedback to the
fuel surface of a pool re in an enclosure, Fire Safety J. 60 (2013) 5663, http://
dx.doi.org/10.1016/j.resaf.2012.12.005.

[2] V. Alireza, S.N. David, W.K. Larry, Effects of altering the liquid phase boundary
conditions of methanol pool res, Exp. Therm. Fluid Sci. 44 (2013) 786791,
http://dx.doi.org/10.1016/j.expthermusci.2012.09.023.
[3] Philippe Dagaut, Abderrahman El Bakali, Alain Ristoridoi, The combustion of
kerosene: experimental results and kinetic modelling using 1- to 3-component
surrogate model fuels, Fuel 85 (2006) 944956, http://dx.doi.org/10.1016/
j.fuel.2005.10.008.
[4] D.D. Benjamin, L. John, K.B. Thomas, C. Marcos, G.B.J. Robert, B.D. Sergey, Pool
res-an empirical correlation, Combust. Flame 160 (2013) 29642974, http://
dx.doi.org/10.1016/j.combustame.2013.06.020.
[5] V. Novozhilov, H. Koseki, CFD prediction of pool re burning rates and ame
feedback, Combust. Sci. Technol. 176 (2004) 12831307, http://dx.doi.org/
10.1080/00102200490457484.
[6] S. Sudheer, S.V. Prabhu, Measurement of ame emissivity of gasoline pool res,
Nucl. Eng. Des. 240 (2010) 34743480, http://dx.doi.org/10.1016/
j.nucengdes.2010.04.043.
[7] B. Chen, S.X. Lu, C.H. Li, Q.S. Kang, V. Lecoustre, Initial fuel temperature effects
on burning rate of pool re, J. Hazard. Mater. 188 (2011) 369374, http://
dx.doi.org/10.1016/j.jhazmat.2011.01.122.
[8] J.M. Chatris, J. Quintela, J. Folch, E. Planas, J. Arnaldos, J. Casal, Experimental
study of burning rate in hydrocarbon pool res, Combust. Flame 126 (2001)
13731383, http://dx.doi.org/10.1016/S0010-2180(01)00262-0.
[9] J.A.R. Woods, B.A. Fleck, L.W. Kostiuk, Effects of transverse air ow on burning
rates of rectangular methanol pool res, Combust. Flame 146 (2006) 379390,
http://dx.doi.org/10.1016/j.combustame.2006.02.007.
[10] J.S. Roh, H.S. Ryou, D.H. Kim, W.S. Jung, Y.J. Jang, Critical velocity and burning
rate in pool re during longitudinal ventilation, Tunn. Undergr. Sp. Technol. 22
(2007) 262271, http://dx.doi.org/10.1016/j.tust.2006.08.003.
[11] J.S. Roh, S.S. Yang, H.S. Ryou, Tunnel res: experiments on critical velocity and
burning rate in pool re during longitudinal ventilation, J. Fire Sci. 25 (2007)
161176, http://dx.doi.org/10.1177/0734904107067300.
[12] J.S. Roh, S.S. Yang, H.S. Ryou, M.O. Yoon, Y.T. Jeong, An experimental study on
the effect of ventilation velocity on burning rate in tunnel res-heptane pool
re case, Build. Environ. 43 (2008) 12251231, http://dx.doi.org/10.1016/
j.buildenv.2007.03.007.
[13] M. Boboulos, M.R.I. Purvis, Wind and slope effects on ROS during the re
propagation in East-Mediterranean pine forest litter, Fire Safety J. 44 (2009)
764769, http://dx.doi.org/10.1016/j.resaf.2009.03.006.
[14] Y. Perez, E. Pastor, A. Agueda, E. Planas, Effect of wind and slope when scaling
the forest res rate of spread of laboratory experiments, Fire Technol. 47
(2011) 475489, http://dx.doi.org/10.1007/s10694-010-0168-7.
[15] C.F. Tao, X.S. Wang, X.N. Zhang, Effects of transverse air ow on mass loss rate
of alcohol pool res in an inclined wind tunnel, J. Civ. Eng. Manage. (2014),
http://dx.doi.org/10.3846/13923730.2014.893920.
[16] C.F. Tao, Y.P. He, Y. Li, X.S. Wang, Effects of oblique air ow on burning rates of
square ethanol pool res, J. Hazard. Mater. 260 (2013) 552562, http://
dx.doi.org/10.1016/j.jhazmat.2013.06.015.
[17] H.M. Chou, K.L. Wong, Heat transfer characteristics of an insulated regular
polygonal pipe by using a wedge thermal resistance model, Energy Convers.
Manage. 44 (2003) 629645, http://dx.doi.org/10.1016/S0196-8904(02)
00064-X.
[18] K.L. Wong, H.M. Chou, Y.H. Li, Complete heat transfer solutions of an insulated
regular polygonal pipe by using a PWTR model, Energy Convers. Manage. 45
(2004) 17051724, http://dx.doi.org/10.1016/j.enconman.2003.09.025.
[19] R.O. Carvel, A.N. Beard, P.W. Jowitt, D.D. Drysdale, Variation of heat release rate
with forced longitudinal ventilation for vehicle res in tunnels, Fire Safety J. 36
(2001) 569596, http://dx.doi.org/10.1016/S0379-7112(01)00010-8.
[20] L.H. Hu, S. Liu, L. Wu, Flame radiation feedback to fuel surface in medium
ethanol and heptane pool res with cross air ow, Combust. Flame 160 (2013)
295306, http://dx.doi.org/10.1016/j. combust ame.2012.10.016.