Page 265 - Dust Explosions in the Process Industries
P. 265

Generation of Explosible Dust Clouds  237


                 Hwang, Singerand Hartz (1974) performed theoretical studies of the dispersion of dust
               in a turbulent gas flow in a duct, following the initial entrainment of the dust deposits
               from the duct wall. In particular, they studied the entrainment of deposited dust by the
               nonstationary turbulent air blast ahead of  a self-sustained dust explosion sweeping
               through a long duct. The objective was to predict the dust concentration in the gas flow
               as a function of time and location in the duct.
                 The dust flux leaving the duct walls was treated as originating from single or multiple
               stationaryor moving sources.Formulas and samplecomputationsfor various types of dust
               somces in circular and rectangular channels were derived based on experimental dust
               entrainment rates. The theoretical results appeared to agree with the physical characteris-
               tics of explosion-drivendust dispersion in a 0.6 m diameter and 50 m long explosiontunnel.
                 In the theoretical analysis, the process of turbulent mixing was treated as a diffusion
               process, using diffusion-type equations that had been successfully applied to the dis-
               persion of  dusts in pipes, open channels, and semi-infinite systems. The generalized
               form of the diffusion equation used was

               ac  -
               -+v  .grad  c = div (k grad c)                                         (3.40)
                at
               where c is the dust concentration,kis the diffusioncoefficient (assumed to be 25-100  cm2/s),
               and  ir is the velocity with which the dust particles were convected, in addition to being
               diffused, V  differs from the gas velocity because of the inertia of the dust particles in the
               flow. It was assumed that the effect of gravity could be neglected during the initial period
               of the dispersion process and that equation (3.40), employing an appropriate value of k,
               determined the gross behavior of the dust cloud.
                 Figure 3.28 shows an example of the computational results obtained. Dust concen-
               trations that would be in the middle of the explosible range for combustible dusts have
               developed at 2.5 m downstream of the dust source. However, at 3.5 m downstream, the
               concentrations are below the typical minimum explosible limit range.
                 Hinze (1975) discussed the Tchen theory of diffusion of discrete solid particles in a
               fluid of homogeneous turbulence. This theory makes the following assumptions:
               1 e  The turbulence of the fluid is homogeneous and steady.
               2. The domain of turbulence is infinite in extent.
               3. The particle is spherical and so small that its motion relative to the ambient fluid fol-
                 lows Stokes’law of resistance.
               4. The particle is small compared with the smallest structure present in the turbulence.
               5. The particle is embedded in the same fluid element during the motion.
               4. Any external force acting on the particle originates from a potential field, such as the
                  field of gravity.

                 All assumptions, except number 5, may in reality actually be satisfied. However, the
               mechanism of a real turbulence is such that it is hardly possible for assumption 5 to be
               satisfied. If the element of fluid containing a small discrete particle could be considered
               nondefonnable, it might satisfy this assumption, provided its size was larger than the
               amplitude of the motion of the discrete particle relative to the fluid (no overshooting).
               However, in turbulent motion, the fluid elements are distorted and stretched into long,
               thin ribbons and it seems; unreasonable that the fluid element should continue to con-
               tain the same discrete particles during this stretching process.
   260   261   262   263   264   265   266   267   268   269   270