Page 266 - Dust Explosions in the Process Industries
P. 266

238  Dust Explosions in the Process Industries


                50,





            m- E
            kl
            I
            c
            .-
            0
            +
            m
            +
            L
            c    01          I          I          I
            (u
            U c
            8  1000,
                   I
            c
            v)                         2.5  rn  downstream of
            2
            U                          dust  source
            d
            m
            U
            0
            -1
                             I          I          I
                  0         0.05       0.10       0.15       0.20
                        Vertical distance from wind  tunnel  floor  Iml

            Figure 3.28  Computed two-dimensionaldust concentration distributionsat two locations in a wind
            tunnel of square cross section 0.53 m x 0.53 m at  1.O s after onset of dust dispersion. Dust source
            is  140 g of rock dust distributed as  a 0.2 m long, even layer over the entire channel width. The aver-
            age wind velocity is 5 m/s and the dust diffusion coefficient k= 50 cm2/s (From Hwang et a/., 1974).

              As part of an account on the use of laser-doppler anemometry to characterize turbu-
            lence, Durst, Melling, and Whitelaw (1981) also discussed various theories for the move-
            ment of small particles in a turbulent flow.
              During the 1980s,a number of furtherexperimental and theoretical studies on the inter-
            action of dust particles and a gas in turbulent flowswere published. Some centralpapers
            are those by Alquier, Gruat, and Valentian (1979); Tomita et al. (1980); Genchev and
            Karpuzov (1980); Tadmor and Zur (1981); Ebert (1983); Elghobashi and Rzk (1983);
            Chen and Wood (1983); Beer, Chomiak, and Smoot (1984); Lee (1984); Krol and Ebert
            (1985); Picart, Berlemont, and Gouesbet (1986); Bachalo, Rudoff, and Houser (1987);
            Johansen (1987); Shrayber (1988); and Lee (1989). These and other similar investiga-
            tions are important to the development of comprehensive computer codes for numeri-
            cal simulation of combustion and explosion of dust clouds (see Chapter 4).
              A number of different methods are now availablefor experimentalinvestigationof the
            turbulence  in gases and dust clouds.  Some of  those  discussed by  Smolyakov and
            Tkachenko (1983) are
              Hot-wire and hot-film anemometer.
              Laser-doppler anemometer.
              Flow visualization by means of small particles (4pm) as “markers.”
              Flow analysis by thermal markers (rapid heating of a small gas volume by hot wires:
   261   262   263   264   265   266   267   268   269   270   271