Page 641 - Dust Explosions in the Process Industries

608 Dust Explosions in the Process Industries

was studied,,and various detonation wave structures were identified. Paplinski and
Wlodarczyk (1994) analyzed the critical conditions for direct initiation of detonations
in dust clouds of infinite size. Klemens et al. (1993) performed experimentsin which det-
onation waves in hybrid mixtures of methane, air, and oats dust were studied. Tulis et al.
(1996), Carve1et al. (1996), and Ven, Olivier, and Gronig (1996) presented experimen-
tal results from different studies of multiple or double fronts in dust cloud detonations
and the dynamic structuralresponse of a tube during dust cloud detonation inside the tube.
Fedorov, Fomin, and Khmel(l996) analyzedreal detonation waves in aluminudoxygen
mixtures mathematically,whereas Fedorov et al. (1998) developed a mathematicalmodel
for steady, self-sustained nonideal detonation of clouds of aluminumparticles in air.These
last workers were able to confirm the existence of steady Chapman-Jouguetdetonation
regimes.Klammer et al. (1999) applied the viscous laminar Navier-Stokesmodel in their
model of generating a dust cloud quasi-detonation inside a plane channel by two dif-
ferent mechanisms. In the first case, the dust was initially deposited as a layer on the
channel floor, and dust dispersion and cloud ignition occurred via a supersonicflow into
the channel. In the second case, the dust was predispersed throughout the channel
volume, and ignition occurred at the heated closed end of the channel. In their theoretical
study of detonation processes in clouds of starch particles in nitrogedoxygen and
hydrogedoxygen atmospheres,Veyssikreet al. (1999) applied a model based on the same
main assumptions as had been used previously to model nonideal detonations of alu-
minum dust dispersed in explosive gadoxygen atmospheres. It was assumed that the
starch particles are gasified by pyrolysis after the temperaturehas reached some critical
value and that the burning rate is controlled by the gasification rate. The model predic-
tions suggested that discrepanciesbetween some earlier experimentalresults were caused
by different particle sizes of the starches and different lengths and diameters of the shock
tubes used in the various experiments.
Klemens et al. (2001b) used similar numerical models to simulate central processes
related to dust explosions in coal mines, including dust layer entrainment behind
a shock wave and initiation of dust cloud deflagration and detonation by alternative
mechanisms.
Zhang, Gronig, and Ven (2001) summarized the extensive work on DDT and stable
detonation waves in dust clouds in air conducted in the Stosswellenlabor of RWTH
Aachen in Germany. The DDT process in long tubes is composed of a reaction com-
pression stage followed by a reaction shock stage as the predetonation process. The
transverse waves that couple the shock wave and the chemical energy release are respon-
sible for the propagation of a stable dust/air detonation. The minimum tube diameters
for DDT and subsequent propagation of stable detonation waves in clouds of most in-
dustrial and agriculturaldusts in air are in the range 0.1-1 .O m and minimum length-to-
diameter ratios for DDT to occur are larger than 100, even when a quite strong ignition
source is used.

9.2.4.1 0
Miscellaneous

Huang, PU,and Ding (1994) observed that burning clouds of aluminumdust in air are elec-
trically conductive.They attributedthis effect to generation of metal vapor by evaporation