Modeling of heat transfer and reactive chemistry for particles in gas-solid flow utilizing continuum-discrete methodology (CDM)

Abstract

Modeling of heat transfer and reactive chemistry for particles in gas-solid flow utilizing continuum-discrete methodology (CDM) by Jordan M. H. Musser A comprehensive multi-phase flow model requires coupled hydrodynamics, boundary conditions, heat and mass transfer, and chemical reaction kinetics. A model must also capture the multi-scale nature of these problems. Computational fluid dynamics-discrete element method (CFD-DEM) provides an accurate description of chemical reactions and heat and mass transfer at the particle scale. Currently, MFIX-DEM, the existing CFD-DEM used as the foundation for this work, can only model coupled hydrodynamics. This dissertation extends the functionality of MFIX-DEM by addressing the remaining deficiencies in three separate efforts. The first effort outlined in this dissertation focuses on the algorithmic development of discrete mass inflow and outflow boundary conditions. This approach allows for the construction of more dynamic models of gas-solid systems. It permits the amount and type of particles to fluctuate during a simulation. Examples illustrating the added functionality are provided. The second investigation explores the three modes of heat transfer in gas-solids systems. Models for particle-particle contact conduction, particle-fluid-particle conduction, particle-gas convection, and particle-particle radiation are selected. Model selection is based on model simplicity, acceptance in existing CFD-DEM heat transfer models, extendibility to particles of different sizes, and computational expense. Modifications are made to selected models before implementing them into MFIX-DEM. The implementation of each model is verified for simple two particle test cases, or in the case of gas-particle convection, a single fixed particle in a flowing gas. Strong agreement is observed between the simulation data and the analytic or numerical solution. Finally, a mathematical interface for managing user-defined particle-gas chemical reactions is developed. The shrinking, unreacted core model is selected as the particle reaction model for its accurate physical account of particle-gas reactions and ability to allow particles to initially contain inert material. The implementation of the reactive chemistry interface is verified for a single reacting particle. Strong agreement is observed between simulation data and the analytic solutions for the particle's mass, species mass fraction, and internal energy equations. Agreement between the simulation data and analytic solution for the shrinking, unreacted core is considered acceptable. ! Update one over the particle's moment of inertia OMOI(NP) = 5.0d0 / (2.0d0 * PMASS(NP) * DES_RADIUS(NP)**2) ! Clear the necessary variables.

Use of Detailed Particle Melt Modeling to Calculate Effective Melt Properties for Powders

“…In order to calculate the temperature-melt fraction relationship, a powder bed is first generated using the discrete element method implemented in the open source software MFIX. In the DEM, particles are modeled as spheres, each with a position, radius, and velocity [8,9]. Particle packings are created by inserting a chosen number of particles in a domain and allowing them to interact with other particles and respond to gravitational forces.…”

Extension of a coarse grained particle method to simulate heat transfer in fluidized beds

“…Instead of using the maximum contact area, Batchelor and O'Brien [27] use the instantaneous normal contact force to calculate the contact radius and the heat transfer is integrated during each time-step. This model is generalized further by using the geometric configuration of the two particles to obtain the contact area and has been implemented in MFIX-DEM [28]. For particle-fluidparticle and particle-fluid-wall conduction, Cheng, Yu and Zulli [29] developed two structurebased analytic models by assuming heat transfer is restricted to the region of a double tapered cone.…”

“…For particle-wall contact conduction term, the model proposed by Batchelor and O'Brien [27] is adopted here. When a particle is in contact with wall, the conduction heat transfer can be calculated as (28) Where r c is the contact radius, k pi and k w are thermal conductive coefficients of particle and wall, T w and T i are the temperatures of wall and particle, and r c can be calculated by the following equation (29) For original system and coarse grained system, r c can be written as:…”

Enhancing the physical modeling capability of open-source MFIX-DEM software for handling particle size polydispersity: Implementation and validation

“…Since its initial release, MFIX has been successfully employed to model a variety of multiphase flow problems relevant to industrial and energy-related applications [13,14,15,16,17,18,19,20,21]. Due to the inherent complexity and diverse range of spatial/temporal scales involved in multiphase flow simulations, several initiatives were carried out to reduce the time-to-solution through performance improvements [22,23,24,25,26,27,28,29,30].…”