Multi-Scale Simulation of Reactive Flow Through a Fixed Bed of Catalyst Particles

Abstract

We investigate the reactive flow through a fixed bed of catalyst particles. Our simulation method belongs to the class of multi-scale approaches and combines nano scale DFT (Density Functional Theory) computations, mean-field kinetics at the micro scale and fully-resolved hydrodynamic simulations at the millimetric scale. At this stage, the simulations at the chemical (nano and micro) scales provide the rate constants needed by the hydrodynamic model in a one-way coupling. In this paper, we focus on the fully-resolved millimetric scale simulation model based on a Distributed Lagrange Multiplier/Fictitious Domain (DLM/FD) method combined with a Finite Volume/Staggered Grid scheme and a high order reconstruction at the fluid/solid interface. In particular, we present various DLM/FD algorithms that supply a divergence free velocity field, a property crucial for the accurate discretization of the additional diffusion-convection equations modeling the chemical reactions. Our method is implemented in our fully MPI platform PeliGRIFF (www.peligriff.com) dedicated to the multi-scale simulation of flows with particles/bubbles/drops. We illustrate the assets of our simulation method on the reactive flow through a pack of 3D cylinders and discuss the forthcoming developments required to extend our approach to a two-way coupling in which the reaction rates are affected by the flow dynamics.

Predicting Average Void Fraction and Void Fraction Uncertainty in Fixed Beds of Poly Lobed Particles

“…This is an extension of the work presented in Dorai et al 13 PRS has been shown to be easily extended to heat and mass transfer as the technique used to solve the momentum equations with solid obstacles, as, e.g., Immersed Boundary, Distributed Lagrange Multiplier/Fictitious Domain or lattice-Boltzmann, can be similarly used to solve other conservation equations. [34][35][36][37] An integrated framework based on DEM and PRS for reactive flows, as demonstrated by Partopour et al 34 and other groups, is currently at the final stage of development in our group 5 and will enable us to assess the chemical efficiency uncertainty induced by the multiple input parameters: loading procedure, particle shape, contact force parameters, dimensionless momentum transfer numbers (Reynolds number), dimensionless mass transfer numbers (fluid/solid diffusivity ratio, Schmidt number, Damkohler number) and kinetic models.…”