Page 429 - Dust Explosions in the Process Industries

396 Dust Explosions in the Process Industries

Table 5.4 Heat conductivities of deposits of some combustible powders and dusts determined
from measurements in a hot wire cell, using equation (5.12)

Source: john and Hensel, 1989.

and A, and Agare the heat conductivity for the solid and gas, respectively,and E is the poros-
ity of the powder deposit (see Chapter 3). As long as A, >> Ag,equation (5.13)reduces to

a=A,(I - + (5.14)
If this equation is applied to Selle’s data in Table 5.3 for powdered sugar, the heat
conductivity becomes 0.70 kJ/mhK, and for aluminum and sulfur,58 and 0.23 kJ/mhK,
respectively.All these values are considerablyhigher than those given by Selle. For cork
dust of porosity 0.95, assuming a value of 2.2 kJ/mhK for A, (same as for sugar), equa-
tion (5.14) yields the value 0.074 kJ/mhK, which is lower than for air and thereforemust
be wrong. The reason is that the simplified equation (5.14) yields A = 0 for E = 1,
whereas according to physical reality a =Ag.This requirement is satisfiedby the more-
comprehensive equation (5.13), which, when applied to the cork data, yields a value of
0.16 kJ/mhK. This differsby only a factor of 2 from the experimentalvalue reported for
cork dust by John and Hensel (Table 5.4).If John and Hensel worked with a significantly
lower porosity than 0.95, this could explain the difference.
Liang and Tanaka (1987b) used the following formula to account for the influence of
temperature on the heat conductivity of cork dust:
a = 6.45~4T+ 0.1589 (w~K) (5.15)

For T = 300 K, this gives A = 0.35 kJ/mhK, which is close to the experimental value in
Table 5.4. For T = 500 K, equation (5.15) gives A = 0.48 kJ/mhK.
Duncan, Peterson, and Fletcher (1988) reviewed various theories for the heat con-
ductivity of beds of spherical particles and compared predicted values with their own
experimental results for 2.38 mm diameter spheres. They found that none of the theo-
ries tested was fully adequate. In particular, the experiments revealed that gas conduc-
tion in the pores between the particles had a significant and predictable effect on the bed
conductivity. For a loosely packed bed of aluminum spheres, the experimentalheat con-
ductivity was 20 and 9 kJ/mhK in nitrogen at atmospheric pressure and in a vacuum,
respectively. For aluminum and a porosity E of 0.35, equation (5.14) yields a bed con-
ductivity of about 400 kJ/mhK, which exceeds the experimental values substantially.
Duncan et al. found that the heat conductivity of beds of aluminum spheres in nitro-
gen increased by a factor of 1.5-2.0 when the bed was exposed to a compacting pres-
sure of about 1 MPa. This effect, which was practically absent in beds of spheres of
nonductile materials, is probably due to enlargement of the contact areas between the
particles in the bed by plastic deformation.