The focus of this thesis is the development of physical models and computational algorithms for the modeling of full-scale catalytic monolithic reactors. Surface reaction is the cornerstone of the operation of such devices, and diffusion is the only mode, locally, by which reactants are transported to the reacting surfaces. In the first part of the study, detailed numerical studies are performed for a representative channel of a typical monolithic reactor to explore the impact of different diffusion models on the simulation results. An optimum diffusion model, henceforth referred to as the Schmidt Number model, is identified for modeling of catalytic reactor operations, based on both accuracy as well as computational efficiency standpoints.
In the second part of the thesis, a new low-memory solver for implicit coupled solution of the species conservation equations is developed. This solver, henceforth known as the IDD+GMRES solver, significantly enhances the stability and convergence of steady state CFD simulations, in comparison to the widely used segregated solution approach. The efficacy of this method is demonstrated using various test cases ranging from pure multi-component diffusion, homogenous combustion of hydrocarbons (laminar flames) and catalytic combustion in both two-dimensional and complex three-dimensional geometries. Use of the IDD+GMRES solver leads to up to 2.5 times reduction in overall computational time and up to 5 times reduction in the time taken by the solution of the species conservation equations, as compared to point implicit Block-Gauss Siedel solvers.
In the third part of the thesis, acceleration of surface chemistry calculation is performed by adapting the In Situ Adaptive Tabulation (ISAT) algorithm for heterogeneous reactions. The ISAT algorithm for surface chemistry is developed from ground up and linked with the CFD code developed in the second part of the thesis. The use of the new ISAT algorithm is demonstrated for channel-scale modeling of catalytic combustion and three-way catalytic conversion and for a full-scale monolithic reactor model of catalytic combustion application. The use of the ISAT algorithm leads to an additional reduction in overall computational time between 50 percent and 150 percent for the cases studied, while speeding-up the surface chemistry calculations alone by up to 11 times.
In the final step, developments resulting from both physical modeling studies and computational algorithm studies are integrated to perform modeling of full-scale monolithic reactors with complex chemistry.
Key contributions of this thesis include investigation and identification of a diffusion model for catalytic monolithic reactor calculations, development of a new low-memory coupled implicit solver for the species conservation equations, the first reported study of the adaptation and use of the ISAT algorithm for large-scale CFD calculations with complex surface chemistry, and integration of all of the above models/algorithms into a single simulation tool for the simulation of full-scale catalytic monolithic reactors.