Page 400 - Dust Explosions in the Process Industries
P. 400

Propagation of Flames in Dust Clouds  369


                 The experimental arrangement consisted of a 20 m3ignition chamber connected to a
               42 m long straight test duct of diameter 0.6 m, which was essentially open at the down-
               stream end. Air was blown through the system at a rate giving 20-30  m/s in the duct,
               and dust was fed into the air streamjust upstream of the 20 m3chamber to give the desired
               dust concentration,ranging from 30 g/m3to 850 g/m3,in the explosion chamber and the
               42 m long duct. The dust cloud was ignited in the 20 m3chamber by a flamejet or a chem-
               ical ignitor. The main results are summarizedin Figure 4.69.




                       ARTINGSTALL  THEORY  1
                                 \/   0    0
                                @I
                                            V C - J  DETONATION
                  25
                                    I           IARTINGSTALLI

               L
               E!
                x
                  15
                                                            THEORY
                  10


                  5

                  0
                              MAXIMUM FLAME SPEED Im/sl

               Figure 4.69  Maximum explosion pressure versus maximum flame speed during coal dust/air explo-
               sions ~IJa 42 m long duct of 0.6 m diameter. Particle size is 87% (mass) 171 pm: m U.S.subbitumi-
               nous coal; o U.K. coal (From Cardner et al.,  1986).


                 Figure 4.69 dso gives the theoreticalrelationshipobtainedbyArtingstall(1961) in solv-
               ing the conservation equations for a steady-statecoal dust/air deflagration. The experi-
               mental relationship found by Bartknecht (1971, 1978)is also included. Gardner et al.’s
               results are in good agreement with Artingstall’s deflagration theory, whereas, on aver-
               age, the Chapman-Jouguetdetonation pressure calculated by Artingstall is significantly
               lower than Gardner et al.’sexperimentalpressures at the calculatedC-J velocity of about
               2350 ds (see Section4.5.3). The extreme experimentalpeak pressure value of 81 bar(g)
               is remarkable. However, Gardner et al. refer to Bull’s argument that, at the onset of det-
               onation,there is always a regime in which the combustion wave is overdriven before set-
               tling down to the C-J conditions. During this transient period, the detonation pressure
               can exceed the C-J value considerably.
                 Gardner et al.’s contribution supports the view that proper detonations can also oc-
               cur in dust clouds and such detonations can be brought about by in situ transition of fast
               deflagrations to detonations via turbulent flame acceleration (DDT, deflagration-to-
               detonation transition), as in premixed gases.
   395   396   397   398   399   400   401   402   403   404   405