Research Sponsor:
Society of Fire Protection Engineers
Educational and Scientific Research
Foundation
Fire Spread
Many fire-spread configurations are not purely upright, but rather inclined at some
angle from the vertical such as wildfires and ceiling fires (Figure 1). The objective of
this study is to investigate the influence of orientation on heat flux profiles in the
combusting plume and subsequent relation to spread rates. Quantifying the effect
of flame spread on inclination angles will improve the accuracy of upward flame
spread models.
g
m f
[4]
[4]
[5]
m f
Vp
xf
[6]
q p
xp
Slowest
spread
(45o & 60o)
Theory [3]
yf
Theory [3]
t ~ xn
ff ((xx,,tt))
qq
f
y m
Fastest spread
(0o, 30o & -30o)
xf
Flame height
xp
Pyrolysis height
yf
q p
Vp
t ~ x
Experimental Approach
A thermally-thick plastic fuel, Polymethyl Methacrylate (PMMA) was ignited evenly
at the base and flames allowed to spread upwards (Figure 2). Seven
thermocouples were melted onto the surface of the fuel to detect the position of the
burning front and eleven thin-skin calorimeters [2] were positioned above the fuel to
measure the heat flux from the flame ahead of the burning surface. The entire
apparatus was placed atop a load cell to measure the rate of mass loss.
Mass-Loss Rates
The mass-loss rate per unit area, m
f normalized by the value at 0 is shown as a
function of angle in Figure 6. Results from the spreading 20 cm long PMMA test in
this study and a 65 cm gas-burner test with sidewalls
from literature [8] exhibit an
[3]
increasing trend with . This is in contrast with [7]two previous theoretical models [3,7]
f varies
and tests with a steadily-burning 5x5 cm square PMMA sample that show m
parabolically with , similar to results for Vp. As [8]> 0, more excess pyrolyzate burns
directly above the pyrolysis region, increasing heat fluxes (primarily radiant) to the
fuel surface, increasing local burning rates. Three dimensional effects also begin to
occur, increasing the heat flux to the burning surface.
5x5cm PMMA
10x20cm PMMA
[3]
[7]
[8]
Figure 6: (Left) The mass-loss rate per unit area, normalized with the value at = 90 is shown here as a
function of . Larger samples with turbulent flames reveal an increasing burning rate per unit area with increasing
while smaller, laminar flames and previous theories predict a parabolic trend with increasing . (Right) 10x20cm
and 5x5cm PMMA samples at = 60, showing significant entrainment from the sides. The influence of the side
entrainment on the 5x5 cm and 10x20 cm samples on the burning rate is still being investigated.
Conclusions
Heat flux profiles ahead of xp exponentially decay. The slope of the decay
increases with decreasing spread rates.
Figure 3: Images were taken perpendicular to the sample surface as it spread upwards, xp
=10cm. Starting from the left ceiling fire, as the inclination angle is increased, underside
flames transition from long blue, well-mixed laminar flames into increasingly turbulent yellow
flames on the topside that lift from the surface, separating from the boundary layer,
dramatically increasing the flame standoff distance yf.
Experimental Results
Measurements of heat fluxes above the fuel surface (x > xp) in Figure 4 reveal that
the heat flux decays exponentially past xp, with an increasing slope as departs 0.
As the slope of these heat flux profiles increases, rates of upward flame spread (Vp)
decrease accordingly, shown in Figure 5 as a function of the angle of inclination, .
Maximum flame spread rates are observed at angles of inclination between -30
and 0 and minimum rates of spread at 60. At angles where > 0 flames lift away
from the surface (Figure 3), resulting in decreased heat fluxes past xp because heat
flux scales inversely with the flame standoff distance, producing lower spread rates.
1.
2.
3.
4.
5.
6.
7.
8.
References