Page 617 - Dust Explosions in the Process Industries
P. 617

584  Dust Explosions in the Process Industries

            9.2.2.2
            Generation of Dust Clouds in Industrial Processes

            Little fundamental, quantitativeknowledge exists about the ways in which primary dust
            clouds are generated in industrial processes. Some work has been done on entrainment
            of dust particles by turbulent gas flows passing over dust layers and ridges in wind tun-
            nels. However, mathematicalmodels connectingessentialparameters of dust cloud gen-
            eration processes and the states of the resulting dust clouds are scarce. Lightstone and
            Raithby (1998) presented a mathematical model that predicts the motion of particles in
            a turbulent flow. The model solves for the probability density function (pdf) of particle
            velocity and treats the impact of  the velocity pdf  as a diffusion process. Particle con-
            centrations arc, in turn, found from the velocity distributions. Good agreement was
            found between model predictions and experimental data.
              When designing experiments for fundamental studies, one should select configurations
            that can be related to industrial practice. Examples are bulk dust dropped into an airflow
            and entrainmentof dust layers by airflows in channels and galleries with and without flow-
            obstructingobstacles.Valuable impulses for further studies can probably be obtainedfrom
            the extensive research carried out worldwide on pneumatic transport of powders.
              Armour-Celu,Woodhead, and Barnes (1998) developed an electrostaticmeasurement
            technique by  which the charging trends during pneumatic transport  of powders in
            pipelines could be inferred. However, processing the measured signals also seemed to
            reveal information about thejow  pattern of the particles in the pipeline. It was indicated
            that the velocity of the particles closest to the pipe wall might be derived from power
            spectral densities. Also, for a constant suspension density, any variation in particle size
            is reflected in the power spectral density.

            9.2.2.3
            Characterizing Dust Cloud Structures: What Is the "Turbulence"
            of a Dust Cloud?

            Comprehensive characterization of turbulence in homogeneous fluids is an extremely
            demandingtask. So, comprehensivecharacterizationof particle-ladenturbulent flows may
            seem close to the impossible. Nevertheless, much valuable research work in this chal-
            lenging and technologically important field has been conducted, and further significant
            work is in progress. Elghobashi (1994) reviewed the progress made and the challenges
            remaining in numerical modeling of particle-laden turbulent flows. Shirolkar et al. (1996)
            discussed some fundamental issues involved in such modeling, such as the key charac-
            teristics of turbulent flow and the nature of the interaction between turbulence and an
            individual particle. They also discussed some key issues involved in comprehensive
            numerical models of  pulverized-coal  combustion. Gouesbet and Berlemont (1 999)
            reviewed the work carried out, since about 1980 by  a research group in France, on
            Eulerian and Lagrangian approaches for predicting the behavior of discrete particles in
            turbulent flows.
              Zhang and Reese (2001)presented a new two-fluid model for gasholids flows, which
            accounts for the influence of the gas turbulence on the random motion of the particles.
            The model comprises a generalizedkinetic theory as well as a new gas turbulencemod-
            ulation model. Simulation results showed that the influence of  gas turbulence on the
   612   613   614   615   616   617   618   619   620   621   622