. Experimental set-upThe test chamb

. Experimental set-up
The test chamber was 3.6 m (L) 1.5 m (W) 0.6 m (H) with
the walls, ceiling, and floor made of two 1 mm thick galvanized
steel plates sandwiching a 30 mm thick rock wool insulation layer,
with two 1.5 m (W) 0.2 m (H) openings A3 and B3 at bottom of
the test chamber as shown in Fig. 1(a).
3.5 L alcohol or diesel was used as the main fuel in the
£500 mm or £400 mm cast iron pan in the middle of the test
chamber. A 3 mm thick, 140 mm (L) 140 mm (W) synthetic rubber
piece in a £200 mm pan S1 or S2 was used as the target fuel to
be ignited on both sides of the main fuel pan, and the center-tocenter
distance between the main and secondary fuel pans is
400 mm, see Fig. 1(b). The electronic balance AND GX-30K and
METTLER PL4002 were used to measure the mass loss rate of fuels
separately.
TS-30B radiant heat flux meters R1–R4 were installed on both
sides of the secondary fuel pan to measure the radiant heat flux
to the floor in fire tests. Smoke composition concentrations were
measured by three MRU VARIO PLUS gas analyzers. Smoke probes
C1 and C2 were installed 150 mm high above the floor and 600 mm
from the center of the main fuel, and the smoke probe C3 was
150 mm below the top and 550 mm from the end of the test model.
Nomenclature
Q0 heat release of fire (kW)
q00 incident heat flux feedback to the fuel bed from the fire
(kW m2
)
Dhc combustion heat (kJ kg1
)
Dhvap vaporization heat of fuel (kJ kg1
)
Af fuel surface area (m2
)
HV1, HV2 height of the ventilation openings (m)
Cd discharge coefficient
cp heat capacity (kJ kg1 K1
)
h thermal convection coefficient (kW m2 K1
)
Greek symbols
v combustion efficiency
q density (kg m3
)
ag emissivity of smoke
r Stefan–Boltzman constant (kW m2 K4
)
s dimensionless time
h dimensionless temperature
e dimensionless heat transfer parameter in Eqs. (6)–(14)
Subscripts
f fuel bed, fuel
0 initial air
g gas property
v vent property
U upper hot smoke zone
L lower zone
R radiation
W walls
out emission or flux from vent
P. Zhang et al. / Applied Thermal Engineering 103 (2016) 1408–1415 1409
Five thermocouple trees (1-T–5-T), each with five thermocouples,
were placed 50 mm below the top with the third thermocouple
tree 3-T just above the main fuel pan for measuring the temperature
evolution of the main fuel flame. The other four thermocouple
trees measured the longitudinal and transverse development of the
smoke temperature in the confined space. All the thermocouples
were £1 mm K-type inconel sheathed thermocouples recorded at
1 s intervals. The experimental setup is shown in Fig. 1(a).
The ventilation condition was regulated by adjusting the frequency
of the smoke discharge fan in the laboratory. The average
velocity in the vent openings was 0.1 m/s in the natural ventilation
cases and 0.5 m/s under the mechanical smoke discharge surroundings.
The critical thermal criteria affecting the ignition of
the rubber were analyzed. Each test case was repeated at least
three times and the typical critical criteria on the ignition of rubber
are listed in Table 1. The next test was started only after the inside
air temperature in the test chamber was lowered to the initial