Real Fire and Fire Modeling
Real Fire and Fire Modeling
Real Fire and Fire Modeling
KUNIO KAWAGOE
Center for Fire Science and Technology
Science University of Tokyo
2641 Yamasaki, Noda-shi, Chiba 278, Japan
ABSTRACT
The system fire research in Japan was started after World War II.
One of the main objectives was the diagnosis of the strength remaining
in buildings burnt out during the as against a future earthquake. The
fire behavior in compartments and the structural behavior of building in
the fully developed stage were studied by carrying out full scale fire
tests in compartments during the 1950s. Later the objectives became the
study of the flashover in a room and the smoke movement inside a
building. Actual buildings scheduled for demolition were used for
investigation of many items during the 1960s and 1970s. In the first
half of this period the main purpose was the observation of real fire
behavior. In the later half, a comparison between the test and the
simulated results became the main objectives. It can be said that
research on building fire in Japan has advanced through such full scale
fire tests.
Key Hords Real Fire Test, Simulation, Burning Rate, Smoke Behavior
Structural Behavior, Modeling and Design
1.INTRODUCTION
In the last fiscal year, full scale gasoline pool fire inside a
compartment were performed on the campus of the Science University of
Tokyo. The compartment was fully open in the front, which was 15m wide
5m long. The depth of the compartment was 24m. In some the 20 tests, a
surprisin phenom
2 2non could be seen. Soon after the ignition of gasoline
in a 10m or 15m pool near the opening, the gasoline vapor flows out
from the compartment close to the floor surface and a big fire whirl
appeared outside of compartment just as if there was another gasoline
pool at the bottom of the fire whirl. These fire whirl could be seen
only when the compartment was receiving the weak wind from behind. The
Karman vortex in the front of opening might exist, but nobody could
expect the appearance of such a fire whirl. The test results of this
project are reported by Dr. T. Mizuno and Dr. O. Sugawa in the 5(B)
session in this symposium.[1]
2
one after another on the B.R.I campus. After the completion of
structural tests, these structures were used for compartment fire tests.
It was fortunate for us to be able to make more than 20 full scale tests
including a ship cabin during 1950's decade.
3
Structural behavior
In each test the expansion and deflection of the compartment
structure and in some tests the vibration of the structure before and
after the fire were measured [11J.
For the concrete block structures of Japan, it was necessary to
insert the reinforcing bar and to pour in situ concrete inside the hole
of joints and to cast the reinforced concrete beam on the top of
concrete block wall together with the reinforced concrete floor. The
difference of expansion between the concrete block wall and reinforced
concrete beam caused breakage of, the concrete block wall creating big
cracks at the upper part of the corner which behavior could not be seen
in the usual wall furnace tests as shown in Figure 2. From the vibration
tests the considerable reduction of rigidity of the total structure
after the fire could be seen.
U 1000 Figure 1
e, 800
w Estimation of time
g§ 600 temperature curve.
!;i
0::
400
W - - - Calculation
0.. 200 - - Data
:a
w 0 L_'------'_---'_---'-_---'-_-'----i
l- 0 15 30 45 60 75 90 105
TIME (min)
........
~ 1000
'-'
w 800
0::
::::>
l- 600
«
w 400
0::
0..
:a 200
w
l- 0
0 15 30 45 60 75 90 105 120 135 Figure 2
T1ME(min)
Expansion of
0.35 Building.
........ 0.30
....!?
0
'-' 0.25
z
0 0.20
(/)
z 0.15
et
w 0.10
X
0.05
0
0 15 30 45 60 75 90 105 120 135
T1ME(min)
4
Figure 3 Explosive Spalling.
Unexpected phenomenon
Through these full scale fire tests we could obtain a lot of
knowledge unexpected before the test such as the explosive spoiling of
thin concrete panel (Figure 4) or the melting down of the wired glass
window caused by the heat transfer coefficient being higher than that in
the testing furnace.
5
Table 1 Field Fire Tests in Real Building
1966 Tokyo Kaijo Tokyo Fire 2nd floor 1. Smoke movement along a
Building Brigade and 3 ro~ms long corridoe (100 m).
(Marunouchi B.R.I. and 24 m 2. Smoke fills in a big room.
Tokyo)2 F.R.I. 4th £loor 3. Smoke spread to upper
2370 m 24 m floor through an open
7 story 4th f~oor stair.
270m 4. Effectiveness of pressu-
rization in a corridor.
5. Effectiveness of a smoke
tower.
6
3. Structural behavior during
the fire.
4. Remaining structural
strength after fire.
1970 New Car Park New Nippon 1st ~loor 1. Behavior of car fires.
Building Steel Co , , 84 m 2. Stuructural behavior of
(Bare steel Ltd. bare steel frame •.
fra2e)
84 m
1 story
1973 Test House B.H.I. and 1st floor 1. Fire growth in a dwelling
Japan Housing J.H.C. one ~welling 2. Smoke movement along a
Corporation 45 m long corridor
testing Center
(Hachioji
Tokyo) 2
new, 45 m
dwelling with
a long corridor
1973 Office Build- B.R. I. and 2nd f~oor 1. Smoke movement in the
ing, Ministry F.R. I. 227 m buildings.
of Health and 2. Pressurization in a stair
Welfare 3. Comparison between
(Kasumigaseki calculated and measured
Tokyok value
512 m
5 story
1975 Fukoku Seimei Tokyo Fire 2nd floor 2 1. Smoke movement inside
Building Brigade 32 m :t 42 m bUilding under several
(Hibiya 270 m boundary conditions.
Tokyo~ 3rd floor 2. Effectiveness of
1003 m 42 m pressurization.
8 story
7
U.S.A .. Waste lumber of 3 tons was scattered about 40m high and at the
end of room near the fire origin the waste lumber was placed along the
wall. All o~tside windows in the long side were opened. Their total area
was 21.6 m • The flame spread very rapidly across the ceiling. It took
only 70 seconds to spread 18m. Then the explosive flashover inside the
whole room followed. The flame emitted through the windows reached the
top of windows in 2 upper floors, and broke the window pane through
which the fire spread to the upper floor. Unexpected sudden high
radiation intensity from flame outside and the rapid smoke fill inside
the building caused much difficulty in making observations and
measurements. We had thought illusion before the test that the ceiling
material might have its own flame spread rate as shown by the result of
tunnel test, but we noticed that it was the violent upward current of
plume and the adequate air that supported such rapid flame spread.
(b)
100 0 ,....,.--r--'--"---'--"----"'-"---'--r--l
,.-.. 800
o
e, 600
400
UJ
ex: 200 '0
~ 0L...l._,---'--,"_.A---L.._-'---L.._"--..........,
t;{
ex: 10 00 I-:==:;::;::::!=~;:::::--'--""""'-"""--"'-'
UJ 800
0. 600
:2
UJ 400
f- 200 3:00
01.-J..._"---'---J._-'----L.._-'---L.._"---1-J
18 12 o
LENGTH (m)
Ceiling
80 em below ceiling
40 em above floor
8
fire spread was depended on the intensity of fire sourcT/~nd the opening
area. In fact, A1/~/At is the former case was 0.078 m and in the
later case 0.016 m •
9
stay there because of the black smoke [13]. In one of the tests in the
Tokyo Kaijo Kasai Building the black smoke reached the open stair 32m
apart from the fire room in 2 minutes and reached the middle of the
corridors in the upper 3 floors in 5 minutes as shown in Figure 5.[15]
10
excess air, the maximum temperature ancr-easeo I.5VV v ana r i.arne vugevueL-
with black smoke spouted out horizontally from windows like jets. Inside
the building heavy smoke was soon ejected out from the stair up for one
and half minutes. Then gradually the air flow was reduced to 300 m3 /min
so that to find the critical value at which the smoke will not come
beyond the swing door inside the open corridor of 0.76m width 1.9m high.
1.0 3
Calculation
Experimental
- E
....
........ 0.5 5
-....
E
........
III .. . >-
r-
u
.- ::J
. .. -
10 lD
Vi
I
I -- 20
>
0
0 10 20 30
TIME (min)
The test was successful. The people who watched the smoke inside the
building were satisfied by the quick ejection of smoke by the operation
of the fan. But the people who watched the test from outside of the
building were astonished by the long horizontal flame jet with heavy
black smoke suddenly emitted from the windows. As the air supply was
gradually reduced to the critical value the flame length became short,
but the long flame jet gave them the impression that the pressurization
method seemed too dangerous. At present some fire experts still do not
like to use the pressurization system. The first impression is very
important. We have shown that we should operate the fan on the weakest
power and then increase the power to find out a critical value.
Structural behavior
One of tests in the Central Hospital Building was a big room fire
to examine the behavior of reinforced concrete structure conducted by H.
Saito [14J. The fire room the third floor was 11m width 22m long in
2in
which of 12 tons (50kg/m ) wood was scattered. The fire temperature
curve corresponded with about 1.5 hours of the standard one. The
horizontal expansion of the upper floor structure was rather rapid
corresponding to the gas temperature rise of fire room. Just after the
gas temperature reached its maximum, the expansion of the structure also
reached its maximum, after which it decreased slowly to alm~~t zero
after one day. The maximum expansion ratio was 3 - 5x10 which
correspond with that of concrete at 300 - 500 °C. The expansion curve
and the maximum expansion ratio suggested us that the expansion of floor
structure was dominated by the temperature rise in the lower part of the
11
floor slab which broke through the restraint of girders.
The maximum displacement of top the of a column in the width
direction was more than 23 mm and its ratio to column height was 1/150,
which was beyond the allowable one for earthquake resisting design. The
displacement of the end column in the length direction was 66mm at which
the glass panes of the window in the staircase wall attached to the fire
room broke down, due to the deformation of its wall.
The behavior of floor slabs was interesting. The floor slabs of
some bays were warped down, some were warped up as shown in Figure 7.
Many cracks appeared on the floor surface, from which a lot of vapor
spouted out to fill the upper floor, which prevented people from making
measurements.
-2 16 10
·········Q········ ..9··········~··········
11 -54 -67 31
Deformation
of floor (mm ) .
12
develop life safety measures and also to estimate any change in the
development of configuration after a future earthquake. A lot of new
knowledge has been obtained for the fire behavior in dwellings [23].
Also the new tendency of the buildings to have big open spaces such
as the air-dome and the atrium has appeared. Full scale fire tests were
performed in several of such buildings to confirm the escape safety [24J.
6. REFERENCE
1) Mizuno, T., and Sugawa, O.,"Experimental Study on Gasoline Station
Fire" Proceeding of the Second International Symposium on Fire
Safety Science
2) Diagnosis and Reinforcement of Burnt Building, Report of BRl No.2,
(1948) (in Japanese)
3) Diagnosis and Reinforcement Concrete Building Damaged by Fire
Exposure, Concrete Pamphlet No.4 Japan Cement Association, (1955)
(in Japanese)
4) Fujita, T., "Fire in Fire Resisting Building and Fire Protection,
Report of BRl, No.1 J-1 (1954) (in Japanese)
5) Kawagoe, K.,"Fire Behavior in Rooms" Report of BRI No.27 (1958)
6) Sekine, T.,"Room Temperature in Fire of a Fire Resistive Rooms"
Report of BRI No.29, (1959)
7) Thomas, P.H."Rcsearch on Fire Using Models", lnst. of Fire
Engineers Quart, (1961)
8) Thomas, P. H. and Heselden, A.J.M. Fire Research Note, No.923 (CIB
Report No.20) (1972)
9) Yokoi, S., "Study on the Prevention of Fire Spread caused by Hot
Upward Current", Report of BRl No.34 (1960)
13
10) Kawagoe, K., and Sekine, T., "Estimation of Fire-Temperature Cur-ve
in Rooms" BRI Occasional Report No.17, (1964)
11) Kawagoe, K., "Damage of Structure in Full-Size Fires" Report of BR:
No.29, (1959)
12) Fire Tests in Mitsubishi Naka 15 Building, "KASAl" JAFSE, Vo1.12,
No.3 (1962) (in Japanese)
13) Fire Test in Daiun Building "KASAl" JAFSE (in Japanese)
14) Fire Tests Report of the Central Hospital of Japan NationaJ
Railway, Association for Railway Building, (1968) (in Japanese)
15) Fire Tests Report of Tokyo Kaijo Kasai Building, "KASAl" JAFSI
Vol.17, No.3 (1967) (in Japanese)
16) Study on the Fire Protection of the Tall Apartments, Japan Housini
Corporation, 74-363, (1974) (in Japanese)
17) Jin, T., Journal of Fire and Flammability Vol.9 (1978)
18) The Study on the High Rise Building, The Association of Buildini
Contractors (1966) (in Japanese)
19) Wakamatsu, T.,"Smoke Movement in Building Fire-Field Experiment ir
Welfare Ministry Building and Analysis of Sennichi Buildini
Fire", BRI Report Paper No.61 (1975)
20) Wakamatsu, T., "Calculation of Smoke Movement in Buildings", Secone
Report, BRl Research Paper No.46 (1971)
21) Report of Fire Test in Yamaichi Stock Company Building, "KASAl",
JAFSE Vol.6,No.3 (1956) (in Japanese)
22) Report of Fire Test in Akabane New Apartment Building, "KASAl",
JAFSE Vol.12, No.4 (1962) (in Japanese)
23) Tsukagoshi, 1., and ltoigawa, E.,"External Radiation at a Full
Scale Fire Experiment", Proceeding of the First International
Symposium on Fire Safety Science (1985)
14