Page 373 - Dust Explosions in the Process Industries

342 Dust Explosions in the Process Industries

I
I
I
0 I I
0 1000 2000 3000
FAN ROTATION SPEED lR.P.Ml

Figure 4.48 Variation of dust flame speed in a horizontal channel with open ends, with rotational
speed of four fans located in the channel, and 300 g/m3of dried wheat flour in air (From Hayes et al.,
1983).

Referring to the work by Chomiak and Jarosinski (1982) on quenching turbulent gas
flames by turbulence, Hayes et al. (1983) attributed the falloff of flame speed in the region
1500 rpm to 3000 rpm to quenching by excessive turbulence. Turbulent flame quench-
ing occurs when the induction time for the onset of combustion exceeds the character-
istic lifetime of the turbulence eddies, so that an eddy composed of hot combustion
products and unburned fluid dissipates before the unburned gas has become ignited.
Hayes et al. did not discuss whether dust could have been separated out at high fan
speeds in regions of nonrandom circulation flow in the channel (cyclone effect). It was
confirmed, by means of hot-wire anemometry, that the degree of turbulence was pro-
portional to the fan speed. For this reason, Hayes et al. used a fan Reynolds number as
a relative measure of the degree of turbulence in the experimental channel.
Klemens et al. (1988) investigated the influence of turbulence on wood and coal
dudair flame propagation in the laboratory-scale flow loop shown in Figure 4.49.
The flow was first streamlined by being passed through a battery of stator blades
upstream of the measurement section. Turbulence was then induced in the first part of
the measurement section by a number of cylindricalrods or rods of V-profiles, mounted
with their axes perpendicular to the main flow direction. The electric spark ignition
source was located immediatelydownstream of the turbulizing zone, and turbulent flame
propagation was observed in the remaining part of the measurement section.Experiments
were conducted with two types of brown coal, a maize dust, and a wood dust-all dusts
being finer than 75 pm particle size. Figure 4.50 shows the average turbulent burning
velocity for maize dudair in the loop as a function of the average normalized turbulence
intensity.
Klemens et al. (1988) observed that their turbulent maize dust flame had the same char-
acteristicnonhomogeneous structureas observed by Proust and Veyssiere (1988) for tur-
bulent corn starcwair flames in a vertical duct.