Page 261 - Dust Explosions in the Process Industries

Generation of Explosible Dust Clouds 233

Figure 3.25 The nozzle for dispersing agglomer-
ates of cohesive dust particles (Yamamoto and
Suganuma, 7 984).

The distribution of effective particle sizes is shifted systematically toward smaller par-
ticles as the dispersion process in the nozzle gets more effective, that is, as the average
air velocity through the nozzle increases.
The effect is quite dramatic.For an airvelocity in the nozzle of 10.5m/s, the median effcc-
tive particle size is somewhat larger than 10 pm; whereas for velocities in the range
100-150 m/s, it is only 1pm. For the primary dust cloud, which was generated in a way
that would be typical in industry,the median particle size would probably be considerably
larger than 10pm. It can be observed from Figure 3.26 that, at the highest air velocities in
the nozzle, the distribution of the sizes of “effective”particles in the secondary dust cloud
approachedthe size distributionfound in a sedimentationbalance after having dispersedthe
powder in a liquid in a way that would be expected to produce close to perfect dispersion.
Yamamoto and Suganuma arrived at the following empirical relationship for the effi-
ciency of the nozzle dispersion process:

(3.35)
xs
where x, is the effective in-situ median particle diameter determined by the cascade
impactor for the actual secondary dust cloud, and x,is the ultimate median particle size
determinedby the sedimentation balance. The parameter h is defined by
h = 0.44Ap,V/do [J/m3s] (3.36)

where Apo is the pressure drop across the dispersing nozzle, V is the mean air velocity
through the nozzle, and dois the orifice diameter.
It is interesting to compare the results in Figure 3.26 with those of Corn and Stein
(1965),in Figure 3.18, from particle dislodgement experimentsin a narrow wind tunnel
of cross section comparable with those of the smallest nozzles in Figure 3.26. The order
of airvelocities required to dislodgeparticles in the sizerange 1-10 pm in Corn and Stein’s
experiments is the same as required to break up agglomerates of 1-10 pm in the Yamamoto
and Suganuma’s nozzle dispersion experiments. This indicates that the adhesive forces
between particles in an agglomerate,a particle, and a plane substrate are of the same nature