Numerical simulations and modelling of mass transfer through random assemblies of catalyst particles: From dilute to dense reactive particulate regime. (21st September 2020)

• Record Type: Journal Article
• Authors: Sulaiman, Mostafa; Climent, Eric; Wachs, Anthony; Hammouti, Abdelkader
• Is Part Of: Chemical engineering science. Volume 223(2020)
• Publication Date: 2020-09-21
• Subjects: Porous catalyst particles -- Chemical reaction -- Mass transfer -- Direct numerical simulations -- Sherwood number; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2020.115659
• Languages: English
• ISSNs: 0009-2509

Abstract

Graphical abstract: Highlights: We propose a model for the mass transfer coefficient accounting for diffusion and internal first-order chemical reaction. We perform direct numerical simulations (DNS) of flow and transfer around particles. A wide range of Reynolds and Schmidt numbers and Thiele modulus was explored. We discuss the limitations of our reactive Sherwood number model. Abstract: We study mass transfer through random assemblies of fixed spherical catalyst particles experiencing an external convective-diffusive fluid stream. Chemical species are transported through the array and are diffused from fluid to solid phase through particles surface. An internal first order irreversible chemical reaction takes place within the porous catalyst particles. We address the determination of mass transfer coefficient by performing direct numerical simulations with fully internal-external coupling using concentration and flux continuity boundary conditions at the solid-fluid interface. We derive a theoretical prediction of the profiles of cup-mixing concentration, average of mean surface and average of mean volume concentration of the particles along the height of the domain. The model for the dimensionless mass transfer coefficient ('reactive' Sherwood number) is accounting for the five dimensionless parameters that control the physics of the system: the Reynolds number Re, the Schmidt number Sc, the Damköhler number Da, the internal-to-external diffusion coefficient ratio γ and the solid volume fraction α s . We use a coupled Sharp Interface/ Discrete Lagrange Multiplier-Fictitious Domain Method (SIM-DLM/FD), thoroughly validated in our previous study (Sulaiman et al., 2019) to test the accuracy of the model over a wide range of dimensionless parameters and solid volume faction (from dilute α s = 0.1 to dense regime α s = 0.5 ). We show and discuss the limitations of the proposed model.