Analysis of the 20 L sphere standard test for flammable dust implementing CFD and DEM computational approaches

Abstract

In the present work, a CFD study was developed over the 20 L sphere standard test for the characterization of flammable dust. The aim of the study was to analyze parameters of the standard test that could affect the results of the test but are not taken into account such as the turbulence, the particle size and the real concentration at the ignition point. Two different approaches for the simulation were studied. First, an Euler-Lagrange coupled approach was considered for the simulation. The continuous phase was modeled with the DES turbulence model, due to its capacity to integrate two different alternatives that will make the solution more robust. For the discrete phase, a spherical particles lagrangian model was used. The material selected was micrometric wheat starch with a particle density of 610 kg/m3. The pressure of the system, the velocity fluctuation and the particle size distribution at the ignition point were validated with an experimental data set. It was found that both, experimental and numeric approaches presents the same behavior with mean absolute errors of 0.183 bar and 5.56 m/s, for pressure and mean velocity fluctuation in x-direction respectively. It was concluded that the rebound nozzle generates a homogeneous distribution of particles within the domain, except for the bottom-right quarter of the sphere where a considerably Iower particle density is presented, due to the location of the canister. However, it was also found that the concentration at the center of the sphere always remains Iower than the nominal concentration?