Document Type thesis Author Name Troupel, Alexandre Email Address atroupel at wanadoo.fr URN etd-052809-145316 Title CFD study of the intra and inter particles transport phenomena in a fixed-bed reactor Degree MS Department Chemical Engineering Advisors A. G. Dixon, Advisor D. DiBiasio, Department Head Keywords fixed-bed reactors intraparticle diffusion CFD Date of Presentation/Defense 2009-05-28 Availability unrestricted
Actual models for fixed-bed reactor modeling make this assumption that temperature is uniform, or at least symmetric, within the catalytic pellets. However, if this holds true for large beds (tube-to-particle diameter ratio N greater than 10), it appears that for small N tubes (N = 3-10) that wall effects cannot be neglected anymore. A large temperature gradient appears in the near wall region. Hence for a particle at the wall a variation in temperature of up to 50¢ªC was noticed.
This temperature change was investigated, and it has been noticed that the proximity to the wall, but also to a low velocity region could explain a maximum in temperature. Furthermore, species concentration discrepancies were also notice. An adiabatic run was made to show that these were not due to heated wall effects. Instead it appeared that these concentration variations are due to both their proximity to a low flow region and to a confined area. Hence incoming diffusion in these zones appeared to be lower than for the rest of the surface.
We also could notice a strong impact of the flow on the temperature patterns in the near wall regions. Hence in our case, it appeared that the 4 holes geometries allowed a better flow in front the particle at the flow, and therefore better transport phenomena. On the contrary, the full cylinder geometry tend to block the flow, consequently temperature on the wall particles were hotter than what they were with the 4 holes cylinder geometry.
A study of the diffusion within the catalytic particles was also conducted. Hence, the Maxwell-Stefan, the dusty gas and the binary friction models were implemented in Fluent. The goal here is to refine step by step the diffusion model used. First products and reactants molar fluxes were assumed to be proportional. The next step was to compute the actual molar fluxes; however this added one more parameter to converge; that is the diffusion coefficient. Finally the assumption of negligible pressure variation within the pellets was dropped. Unfortunately, the implementation into Fluent was not successful, and few possible reasons were given.
Browse by Author | Browse by Department | Search all available ETDs
Questions? Email firstname.lastname@example.org