Unsteady 3-D RANS simulations of dust explosion in a fan stirred explosion vessel using an open source code

Journal article, 2020

Dust explosion is a constant threat to industries which deal with combustible powders such as woodworking, metal processing, food and feed, pharmaceuticals and additive industries. The current standards regarding dust explosion venting protecting systems, such as EN 14491 (2012) and NFPA 68 (2018), are based on empirical correlations and neglect effects due to complex geometry. Such a simplification may lead to failure in estimating explosion overpressure, thus, increasing risk for injuries and even fatalities at workplaces. Therefore, there is a strong need for a numerical tool for designing explosion protecting systems. This work aims at contributing to the development of such a tool by (i) implementing a premixed turbulent combustion model into OpenFOAM, (ii) verifying the implementation using benchmark analytical solutions, and (iii) validating the numerical platform against experimental data on cornflour dust explosion in a fan-stirred explosion vessel, obtained by Bradley et al. (1989a) under well-controlled laboratory conditions. For this purpose, the so-called Flame Speed Closure model of the influence of turbulence on premixed combustion is adapted and implemented into OpenFOAM. The implementation of the model is verified using exact and approximate analytical solutions for statistically one-dimensional planar and spherical turbulent flames, respectively. The developed numerical platform is applied to unsteady three-dimensional Reynolds Averaged Navier-Stokes simulations of the aforementioned experiments. The results show that the major trends, i.e. (i) a linear increase in an apparent turbulent flame speed St,b with an increase in the root mean square (rms) turbulent velocity u' and (ii) and an increase in St,b with an increase in the mean flame radius, are qualitatively predicted. Furthermore, the measured and computed dependencies of St,b(u') agree quantitatively under conditions of weak and moderate turbulence.