Вы находитесь на странице: 1из 7

Geometric effect of radiative heat exchange in concave structure with application

to heating of steel I-sections in re


Zhi-Hua Wang
*
Department of Civil and Environmental Engineering, Princeton University, NJ 08544, USA
a r t i c l e i n f o
Article history:
Received 7 August 2009
Received in revised form 24 October 2009
Accepted 24 October 2009
Available online 1 December 2009
Keywords:
Radiation
Geometric effect
View factor
Reection
I-section
Fire
a b s t r a c t
In structural re engineering, it is well-known that the radiative heat exchange in unprotected steel I-sec-
tion is attenuated due to geometric effect. An analytical approach to formulate the geometric effect by
employing view factors is presented in this paper for concave structures in general. In particular, the
exact solution is extended to predict the net radiation ux and the temperature evolution for unprotected
steel I-sections in re, which are fully characterized by the I-section aspect (width-to-depth) ratio. The
proposed method is validated by comparison to the conventional model in Eurocode 3, as well as to
experimental investigations. The method predicts the transient temperature eld inside the I-sections
with reasonable accuracy, while Eurocode provision yields over-estimations. Besides, the effective emis-
sivity is found to be insensitive to temperature, but rather controlled by the section aspect ratio.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
For radiative heat transfer in concave structures, thermal
energy emitted from one surface can be obstructed and reected
by other surfaces. The effect of geometry is then a key to under-
standing characteristics of radiation in concave structures. Geo-
metric view factors between surfaces of concave structure need
to be incorporated in order to calculate the thermal energy balance
accurately [1,2]. View factors for specic geometric congurations
were determined in a number of earlier research works (e.g. [3,4]).
In general, the view factors between two surfaces can be evaluated
either using contour integration formula or in discrete manner.
One major application of radiative heat transfer in structure re
engineering is to evaluate the energy balance in steel I-sections sub-
ject to elevated temperature. Radiation is predominant for steel
structure subject to high temperature, for conduction is almost
instantaneous due to high thermal conductivity (45 W/mK) and
canbe neglectedfor giventimescaleof interest. Needless tomention
I-sections are the most commonly used steel members in building
and construction, they are also stereotypical concave structures
where radiation heat transfer is complicated by geometric effect. It
is well-knownthat theprovisions inreengineeringcodes [5,6] tend
to overestimate the temperature eld for bare (unprotected) steel I-
sections. The attenuation in radiation is associated with two main
reasoning. First and the intrinsic reason is that the geometric effect
associated with the concave structure of I-sections, where radiation
are trapped and exchanged by innite (theoretically) reections be-
tween anges and web. Secondly, for I-sections subject to natural
res, the participating media (aerosol, smoke, etc) are in general
non-isothermal, where the effect of absorption, emission and scat-
tering is signicant and cannot be neglected.
The geometric effect of thermal radiation in steel I-section was
rst studied by Wickstrom [7]. A bulk factor called shadow effect
coefcient was proposed to count for the radiation attenuation
due to the presence of void between anges and web in an I-sec-
tion. Later the usage of the bulk coefcient was conceptually clar-
ied in [8] and was adopted by the international standardization
work [5]. But exact solution for the geometric effect, particularly
for I-sections by employing view factors and multiple (innite)
reections, remains hitherto absent. On the other hand, the effect
of non-isothermal participating medium was investigated by
Wong [9] and Ghojel and Wong [10]. A phenomenological mean
beam length was adopted and the total emissivity/absorptivity
of combustion product was formulated, based on polynomial
regression and empirical relation.
Analytical formulation for radiative ux using optic length of
non-isothermal participating medium was given by Heaslet and
Warming [11]. An alternative analytical procedure based on two-
ux method was proposed in [12], where the physical basis of
non-isothermal effect of participating medium was claried. In
general, for unprotected steel concave sections under elevated
temperature, attenuation of radiative heat transfer is a combined
result from the interaction of geometric effect and non-isothermal
0017-9310/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijheatmasstransfer.2009.11.013
* Tel.: +1 609 258 5334; fax: +1 609 258 2760.
E-mail address: zhihuaw@Princeton.EDU
International Journal of Heat and Mass Transfer 53 (2010) 9971003
Contents lists available at ScienceDirect
International Journal of Heat and Mass Transfer
j our nal homepage: www. el sevi er . com/ l ocat e/ i j hmt
effect of participating medium. Despite its relative simplicity as
compared to insulated structures (e.g. [13]), accurate formulation
of radiative energy balance relies on quantitative determination
of both attenuation mechanisms.
In this paper, we focused on the geometric effect on the radia-
tive heat transfer occurred in concave structures. The analytical
procedure for general concave conguration was outlined by Spar-
row and Cess [1]. The exact solution of the geometric effect was ap-
plied to study the diurnal radiation exchange inside an urban street
canyon by Harman et al. [14]. The conguration of steel I-section is
analogous to a street canyon model. In this paper, the analytical
procedure is extended to solve for radiation exchange in I-sections
exposed to re, with different boundary conditions and the param-
eter space from urban canyon model.
2. Diffuse radiative exchange in concave structure
In this section, we follow the analytical procedure rst devel-
oped by Sparrow and Cess [1], which was later applied to radiative
heat transfer in an urban street canyon by Harman et al. [14]. The
general assumptions are as follows. All surfaces are assumed Lam-
bertian. The radiative reections are diffuse and isotropic. Specular
reection is not considered in this study. Participating medium be-
tween solid surfaces is assumed to be isothermal and radiatively
non-absorbing, i.e. the absorption, emission and scattering by air
are neglected.
2.1. Exact solution
Given ith facet in a concave structure, where 1 6 i 6 N and N is
the total number of facets, it is associated with a range of radiative
uxes, namely the total incoming K
i
, the total outgoing B
i
, the total
emitted X
i
and the net radiative ux q
i
, respectively. These uxes
are not independent but related by
K
i

N
j1
F
ij
B
j
1
B
i
X
i
1 e
i
K
i
2
q
i
K
i
B
i
3
where subscripts i and j are facet indices, e
i
is the emissivity and
F
ij
are the view factors for radiation from ith to jth surface. Proper-
ties that must be satised by view factors matrix include: self view
factor for at (convex) facet is zero, and no radiant energy can be
lost, i.e.
F
ii
0; no summation over i;

N
j1
F
i
j
1 4
The quantity of interest is the net radiation ux, which involves the
outgoing radiation from the other facets. Combining Eqs. (1) and (2),
we have
B
i
X
i
1 e
i

N
j1
F
ij
B
j
5
Clearly the solution of the problem involves recurrence of B
i
, the re-
ected portion of radiation incurred at each surface. Exact solution
therefore invokes solving the geometric series associated with mul-
tiple (innite) radiative reections. Rewrite Eq. (5) as
X
i
B
i
1 e
i

N
j1
F
ij
B
j

N
j1
C
ij
B
j
6
where C
ij
= d
ij
(1 e
i
)F
ij
. The matrix C
ij
always has an inverse,
which is denoted as [W
ij
] = [C
ij
]
1
. Thus for each facet:
B
i

N
j1
W
ij
X
j
; K
i

B
i
X
i
1 e
i
7
q
i

N
j1
F
ij
X
j
X
i
if e
i
1
e
i

N
j1
W
ij
X
j
X
i
1e
i
if e
i
1
_

_
8
For each facet, the material emissivity and temperature are known
quantities. The emitted heat ux can then be expressed using Boltz-
manns law:
X
i
e
i
rT
4
i
9
where r 5:67 10
8
W=m
2
K
4
is the Stefan-Boltzmann constant.
Note that Eqs. (7) and (8) are the exact solution of the diffuse radi-
ation exchange in a concave structure with innite reections. In
the subsequent sections, these exact solutions will be used to solve
numerical examples.
Nomenclature
b width of ange of I-section, m
B total outgoing radiative ux, W/m
2
c specic heat, J/kg K
D depth of I-section, m
F geometric view factors
h
c
convective heat transfer coefcient, W/m
2
K
H
p
/A
s
section factor, m
1
q net heat ux, W/m
2
r aspect (width-to-depth) ratio
T
g
re gas temperature, C
T
s
steel temperature, C
Greek letters
e emissivity
g shadow effect coefcient
h absolute temperature, K
q density, kg/m
3
r Stefan-Boltzmann constant, W/m
2
K
4
K total incoming radiative ux, W/m
2
X total emitted radiative ux, W/m
2
Subscripts
b blackbody
c convection
eff effective
g re gas
i, j, k surface indices
p perimeter
r radiation
res resultant
s steel
Superscript
0, 1 order of reection (1 e) terms retained in approximate
solutions
998 Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003
2.2. Approximate solution
Approximate solutions of diffuse radiation exchange in concave
structure can be obtained by ignoring all reection (terms of order
(1 e) and higher) or retaining only one reection. From Eq. (8),
the zeroth and rst order approximate solutions are given by
q
0
i
e
i

N
j1
F
ij
X
j
X
i
10
q
1
i
e
i

N
j1
F
ij
X
j
1e
j

N
k1
F
jk
X
k
_ _
X
i
q
0
i
e
i

N
j1
F
ij
X
j
1e
j

N
k1
F
jk
X
k
_ _
11
respectively, where superscripts 0 and 1 denote the order of
(1 e) terms retained. Note in the case of e
i
= 1.0, i.e. blackbody
radiation, the approximate and exact solutions are identical. Due
care must be exercised by using these approximations. The errors
of the approximate solutions are very sensitive to the material
emissivity. Detailed discussion on comparison of the exact and
approximate solutions can be found in [14].
3. Application to I-sections under re condition
The schematic of a steel I-section exposed to re is shown in
Fig. 1. The concave structure consists of four facets: three steel sur-
faces (anges and web) and one sealing surface of the void by a re
source (e.g. heating panels, furnace, burning objects, etc), with
numbering shown in the gure. Due to symmetry in conguration,
only the upper half of the model is analyzed while that of the lower
part is identical.
3.1. Matrix of view factors
Determination of the view factors is crucial to study of the geo-
metric effect in concave structures. With the surface numbering in
Fig. 1, the view factors can be expressed as [1]:
X
1
F
12
F
21

1 r
2
_
r 12
X
2
F
34
F
43

1 1=r
2
_
1=r 13
X
3
F
13
F
14
F
23
F
24
0:51 F
12
14
X
4
F
31
F
41
F
32
F
42
0:51 F
34
15
where r = b/D is the (width-to-depth) aspect ratio of the I-section. In
matrix form, the view factor matrix F can be written as:
F F
ij

0 X
1
X
3
X
3
X
1
0 X
3
X
3
X
4
X
4
0 X
2
X
4
X
4
X
2
0
_

_
_

_
16
Clearly the matrix satises the properties in Eq. (4). Note that the
matrix is fully characterized by a single variable, viz. the section
aspect ratio r. Dummy indices are used for the numbering of the
facets in Fig. 1, i.e. if the facets are numbered in other manner,
matrix F can be formulated simply by permutation of the elements
in Eq. (16).
A plot of view factors versus aspect ratios is shown in Fig. 2. For
practical aspect ratios of structural I-sections, the view factors can
vary in a wide range. It is shown in Fig. 2 that for X
1
and X
2
, the
view factors between two parallel surfaces, the numerical values
approach the limit of 1.0 if the distance between the two parallel
surfaces are innitesimal. On the other hand, X
3
and X
4
are the
view factors between surfaces in right angle, the maximum values
are 0.5. All view factors reduce to zero if two parallel surfaces are
far apart, or the target surface area is extremely small.
3.2. Steel temperature formulation
In current Eurocode 3 (EC3) [5] for re engineering design, tem-
perature eld inside unprotected steel sections is formulated in an
incremental manner, as:
DT
s
g
H
p
=A
s
c
s
q
s
q
c
q
r
Dt 17
where T is temperature in C, g is a bulk shadow effect coefcient,
H
p
/A
s
is the section factor of an I-section, c is the specic heat, q
is the density, subscript s designates steel, p perimeter, and c
and r convective and radiative heat transfer, respectively. Note
here we use the bulk capacitance assumption for steel, i.e. temper-
ature inside the entire I-section is uniform and equal to the surface
temperature, due to high thermal conductivity.
The convective heat ux q
c
is given by
q
c
h
c
T
g
T
s
18
where h
c
is the convective heat transfer coefcient and T
g
is the
temperatures of re gas. The radiative heat ux q
r
is calculated
using
1
2
3
4
b
D
Fig. 1. Schematic of the radiative heat transfer model for a steel I-section under re.
Section aspect ratio r = b/D
V
i
e
w

f
a
c
t
o
r
s
10
-2
10
-1
10
0
10
1
0
0.2
0.4
0.6
0.8
1
X
1
X
2
X
3
X
4
Fig. 2. View factors for steel I-sections with different aspect ratios.
Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003 999
q
r
e
res
r h
4
g
h
4
s
_ _
19
where e
res
is the resultant (or effective) emissivity, h = T + 273 is the
absolute temperature in K.
It is clear that in Eq. (19), the geometric effect of concave I-sec-
tion is not considered. A bulk coefcient g 6 1.0 is used in Eq. (17)
instead, where no solid physical ground has been laid for the
choice of value of g. For conservative design, EC3 suggests that
g = 1.0 and h
c
= 25 W/m
2
K. It is not surprising that using g = 1.0
consequently results in over-estimation of steel temperature eld.
Some researchers (e.g. Wong and co-workers [9,10]) opted to ad-
just parameters for radiative transfer, recognizing that the partici-
pating medium in real re is not ideally isothermal. Polynomial
regression was proposed based on empirical data for total emissiv-
ity/absorptivity in [9,10].
It was pointed out in [12] that for unprotected steel section
under real re, the participating medium consists of aerosol and
soot, is optically thick. The effect of absorption, emission and
scattering is signicant and the effective emissivity is strongly
dependent on optical length of the participating medium. In con-
trast, for unprotected steel section under standard re condition
controlled in furnace tests, the participating medium is assumed
to be relatively transparent (optically thin). Therefore the geo-
metric effect on radiative heat transfer is dominant. Adopting
the solutions presented in Section 2, we can obtain the net radi-
ative heat ux incident at each surface of an I-section. The effec-
tive net radiative heat ux received by the steel section (along
inner perimeter) can then be expressed in an averaged (spatially)
sense as
q
r;eff

q
web
2rq
flange
1 2r
20
where q
web
and q
ange
are the net radiative heat uxes at web (sur-
face #2 in Fig. 1) and anges (surfaces #3 and #4 in Fig. 1) respec-
tively, which are computed using Eq. (8) in turn. Not that beside the
emissivities and temperatures of individual facets, the net radiative
heat ux is fully determined by the section aspect ratio r. Eq. (20)
gives a measure of how the temperature of the I-section will
change. Substituting Eq. (20) into the steel temperature formula-
tion, and ignoring g, a new steel temperature formulation can be
obtained as
DT
s

H
p
=A
s
c
s
q
s
q
c

q
web
2rq
flange
1 2r
_ _
Dt 21
Note that as comparing to Eq. (19), there is no need to recourse to
an explicit formulation of effective (resultant) emissivity of the
lumped radiative heat transfer model. Instead material emissivity
of steel is used in formulation of the geometric effect in Eq. (9).
An equivalent (effective) emissivity for the radiation process natu-
rally arises as a consequence of incorporation of the geometric ef-
fect, which will be discussed shortly.
4. Results and discussion
The exact solution of radiative heat transfer in concave struc-
ture (Eq. (8)) and its application to unprotected steel structures
in re (Eq. (21)) are presented in this section. We rst validate
the exact solution using a particular case where temperatures are
invariant. An equivalent (over at surface) net radiative heat ux
in the concave structure is dened. Based on the denitions, the
physical meaning of the geometric effect is claried. Lastly, the
proposed method for temperature formulation is validated by
comparison to experimental results.
4.1. Radiation in I-section with constant temperature
Consider the steady state case in which radiative heat transfer
takes place between a heating panel and an I-section (not neces-
sarily steel), both maintained at constant temperatures. The pre-
scribed temperatures for the heating panel and the I-section are
h
g
= 315 K and h
s
= 295 K, respectively. Emissivity for heating
source is set to be 1.0. Results are obtained using the exact solution
for a range of section aspect ratios and material emissivity of I-
sections.
As mentioned earlier, the net radiative heat ux received along
the inner perimeter of the I-section gives a measure of how the
temperature of the section will change. Using Eq. (20), an effective
emissivity can be dened as
e
eff

q
r;eff
q
b

q
web
2rq
flange
1 2rr h
4
g
h
4
s
_ _ 22
where q
b
r h
4
g
h
4
s
_ _
is the net blackbody radiation ux between
two objects with prescribed temperatures. Effective in this sense
means the value the parameter would have to take if the I-section
is at. The equivalent at surface in Eq. (22) would be viewed
as if the inner perimeter of the I-section is stretched to be straight.
Fig. 3 shows the contour plots of the effective emissivity as
function of section aspect ratio and material emissivity. First note
that, as r ?0, the concave I-section reduces to a at surface, and
the effective emissivity convergences to the material emissivity.
It is also shown that for a given material emissivity, value of
parameter e
eff
decreases with an increase in section aspect ratio,
hence reduces the magnitude of net radiation received by the sec-
tion inner perimeter. Physically, it means that the presence of con-
cave geometry results in a radiative attenuation and the
temperature rise is retarded.
4.2. Unprotected steel I-section in standard re
Now consider more realistic cases: temperature prediction of
unprotected steel I-sections subject to standard re. Experimental
results were reported by Wainaman and Kirby [15]. Both temper-
ature formulations, i.e. the one with a bulk shadow coefcient gi-
ven by EC3 (Eq. (17) and the proposed method (Eq. (21)) using
exact solution for geometric effect, are compared against the test
results.
Log(r)
M
a
t
e
r
i
a
l

e
m
i
s
s
i
v
i
t
y

-2.0 -1.0 0.0 1.0


0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.9
0.1
0.3
0.5
0.7
Fig. 3. Effective emissivity e
eff
as function of aspect ratio.
1000 Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003
The standard re temperature is given by
T
g
t T
0
345log
10
8t 1 23
where T
0
is the initial temperature, and t is time (in min).
Four tested specimens from the category called indicative col-
umn tests (with four faces exposed to re) [15] are adopted in this
study, with the section factor of 30, 76, 133 and 149, respectively.
The aspect ratios of all the specimens are around 0.55, which is
typical for most of the steel I-sections used in buildings and struc-
tures. Furnace temperatures were carefully controlled to model the
standard re curve as closely as possible. Experimental tests were
carried out at different seasons, so initial temperature of the tests
varied widely from 0 to 27 C. Numerical parameters used include:
Dt = 30s, e
s
= 0.6, e
g
= 0.8, e
res
= 0.5 (for EC3), c
s
q
s
= 4.71 10
6
J/
m
3
K, g = 1.0 and h
c
= 25W/m
2
K.
Results of comparisons of steel temperatures are shown in Figs.
47. It is clear that EC3 provision over-estimates the steel temper-
ature eld, even when a low resultant emissivity e
res
= 0.5 is used.
In contrast, the agreement between the proposed approach and the
test results are reasonably good, for I-sections with a range of sec-
tion factors and initial temperatures. Numerical experiments also
show (not presented in this paper) that as the aspect ratio r ?0,
results predicted by the proposed approach and EC3 collapse to
the same curve, as expected.
In addition to the steel temperature prole, evolution of the
effective net radiative heat uxes and the effective emissivity are
presented in Figs. 811. It is interesting to note that for the case
of unprotected I-sections in standard re, the effective emissivity
e
eff
is nearly independent of temperature in the rapid (rst
45 min) heating phase, though both the numerator q
r,eff
and the
denominator q
b
in Eq. (22) vary widely with temperature. At latter
testing stage, the effective emissivity varies slightly with a discrep-
ancy less than 10% (see Fig. 11). For all the four cases, values of the
effective emissivity range between 0.27 and 0.30. The insensitive-
ness of e
eff
is due to the fact that the test specimens have similar
values of aspect ratio. On the contrary, earlier studies [9,10] as-
sumed that the effective (total) emissivity for radiative heat trans-
fer in unprotected steel members is a function of temperature only.
Our observation suggests that the section aspect ratio, instead of
temperature of steel, may be the controlling parameter for effec-
tive emissivity due to geometric effect.
5. Concluding remarks
In structural re engineering, rigorous formulation and physical
basis for the geometric effect of radiation occurred in unprotected
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
0 10 20 30 40
0
150
300
450
600
750
900
Test: flange
Test: web
Proposed
EC3
standardfirecurve
Fig. 4. Comparison of steel temperature predictions by the proposed method, EC3
and experimental results (T
0
= 20 C, r = 0.554, H
p
/A
s
= 35 m
1
).
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
0 10 20 30 40
0
150
300
450
600
750
900
Test: flange
Test: web
Proposed
EC3
standard fire curve
Fig. 5. Comparison of steel temperature predictions by the proposed method, EC3
and experimental results (T
0
= = 20 C, r = 0.543, H
p
/A
s
= 76 m
1
).
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
0 5 10 15 20 25 30
0
150
300
450
600
750
900
Test: flange
Test: web
Proposed
EC3
standard fire curve
Fig. 6. Comparison of steel temperature predictions by the proposed method, EC3
and experimental results (T
0
= 16 C, r = 0.544, H
p
/A
s
= 133 m
1
).
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
0 20 40 60 80 100 120
0
200
400
600
800
1000
Test: flange
Test: web
Proposed
EC3
standard fire curve
Fig. 7. Comparison of steel temperature predictions by the proposed method, EC3
and experimental results (T
0
= 27 C, r = 0.541, H
p
/A
s
= 149 m
1
).
Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003 1001
I-sections exposed to re have long been absent. In this study, the
exact solution of geometric effect for radiative heat transfer is pre-
sented, for concave structure in general. The exact solutions is
completely characterized by a controlling parameter, viz. the sec-
tion aspect ratio, r = b/D. As contrary to what was speculated in
earlier studies, the effective emissivity due to geometric effect is
found to be insensitive to temperature from the numerical exper-
iments. Rather, our study suggests that the effective emissivity
may be a strong function of the controlling parameter, i.e. the sec-
tion aspect ratio. The exact solution is extended to predict the tem-
perature evolution of unprotected steel I-sections in standard re,
where the results are in reasonably good agreement with experi-
mental data.
In general, attenuation of radiative heat transfer for I-sections in
re is a combined result of the geometric effect due to concave
structure as well as the effect due to presence of non-isothermal
(absorbing/emitting/scattering) medium. For I-sections, geometric
effect is ubiquitous and inherent. Attenuation effect due to pres-
ence of participating medium, on the other hand, can be quite sig-
nicant in natural re scenarios. Therefore accurate models for
both mechanisms are required in order to yield safe design and
analysis in structural re engineering.
References
[1] E.M. Sparrow, R.D. Chess, Rad. Heat Transfer, Augmented ed., Hemisphere
Publishing Corporation, Washington, 1978. pp. 81144, 339349.
[2] R. Siegel, J.R. Howell, Therm. Rad. Heat Transfer, fourth ed., Taylor & Francis,
New York, 2002. pp. 155192.
[3] A. Ambirajan, S.P. Venkateshan, Accurate determination of diffuse view factors
between planar surfaces, Int. J. Heat Mass Transfer 36 (8) (1993) 22032208.
[4] V.R. Rao, V.M.K. Sastri, Efcient evaluation of diffuse view factors for radiation,
Int. J. Heat Mass Transfer 39 (6) (1996) 12811286.
[5] CEN, prEN 1993-1-2, Eurocode 3: Design of steel structures Part 1-2: General
rules Structural re design, Stage 49 draft, Brussels, 2003.
[6] Society of Fire Protection Engineers (SFPE), SFPE Handbook of Fire Protection
Engineering, third ed., Maryland, 2002.
[7] U. Wickstrom, Calculation of heat transfer to structures exposed to re
shadow effects, in: Interam Proceedings of the Ninth International
Conference, vol. 1, Interscience Communication Ltd., Edinburgh, 2001, pp.
451458.
[8] U. Wickstrom, Heat transfer by radiation and convection in re testing, Fire
Mater. 28 (5) (2004) 411415.
[9] M.B. Wong, Size effect on temperature of structural steel in re, J. Struct. Eng.-
ASCE 131 (1) (2005) 1620.
[10] J.I. Ghojel, M.B. Wong, Heat transfer model for unprotected steel members in a
standard compartment re with participating medium, J. Construct. Steel Res.
61 (6) (2005) 825833.
Time (min)
R
a
d
i
a
t
i
v
e

h
e
a
t

f
l
u
x

(
k
W
/
m
2
)

e
f
f
0 10 20 30 40
0
10
20
30
40
50
60
0.0
0.1
0.2
0.3
0.4
0.5
q
r,eff
q
b

eff
Fig. 9. Evolution of the net radiative heat ux and effective emissivity for I-section
with T
0
= 20 C, r = 0.543, H
p
/A
s
= 76 m
1
.
Time (min)
R
a
d
i
a
t
i
v
e

h
e
a
t

f
l
u
x

(
k
W
/
m
2
)

e
f
f
0 10 20 30 40
0
10
20
30
40
50
0.0
0.1
0.2
0.3
0.4
0.
5
q
r,eff
q
b

eff
Fig. 10. Evolution of the net radiative heat ux and effective emissivity for I-section
with T
0
= 16 C, r = 0.544, H
p
/A
s
= 133 m
1
.
Time (min)
R
a
d
i
a
t
i
v
e

h
e
a
t

f
l
u
x

(
k
W
/
m
2
)

e
f
f
0 20 40 60 80 100 120
0
10
20
30
40
50
0.0
0.1
0.2
0.3
0.4
0
.5
q
r,eff
q
b

eff
Fig. 11. Evolution of the net radiative heat ux and effective emissivity for I-section
with T
0
= 27 C, r = 0.541, H
p
/A
s
= 149 m
1
.
Time (min)
R
a
d
i
a
t
i
v
e

h
e
a
t

f
l
u
x

(
k
W
/
m
2
)

e
f
f
0 10 20 30 40
0
20
40
60
80
100
0.0
0.1
0.2
0.3
0.4
0.5
q
r,eff
q
b

eff
Fig. 8. Evolution of the net radiative heat ux and effective emissivity for I-section
with T
0
= 20 C, r = 0.554, H
p
/A
s
= 35 m
1
.
1002 Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003
[11] M.A. Heaslet, R.F. Warming, Radiative transport and wall temperature slip in
an absorbing planar medium, Int. J. Heat Mass Transfer 8 (7) (1965) 979994.
[12] Z.H. Wang, K.H. Tan, Radiative heat transfer for structure members exposed to
re: an analytical approach, J. Fire Sci. 26 (2) (2008) 133152.
[13] Z.H. Wang, K.H. Tan, Greens function approach for heat conduction:
application to steel members protected by intumescent paint, Num. Heat
Transfer B: Fund. 54 (6) (2008) 435453.
[14] I. Harman, M.J. Best, S.E. Belcher, Radiative exchange in an urban street canyon,
Bound.-Layer Meteorol. 110 (2) (2004) 301316.
[15] D.E. Wainaman, B.R. Kirby, Compendium of UK standard re test data:
Unprotected steel 1, Appendix A, British Steel Corporation, Rotherham, 1998,
data sheet 6982.
Z.-H. Wang / International Journal of Heat and Mass Transfer 53 (2010) 9971003 1003

Вам также может понравиться