(355b) Simulating Fast Chemical Reactions and Mixing with Random Coalescence-Dispersion Modeling

Patterson, G. K. - Presenter, Missouri University of Science and Technology

A general code for simulation of turbulent flow and mixing with fast multiple chemical reactions using a combination of k-ε CFD and random coalescence-dispersion (C-D) is presented.  Previously a code was developed to do approximate simulations of mixing and chemical reaction by utilizing C-D and a simple MATLAB simulation of the flows through volume segments of a stirred reactor vessel.  Flows and rates of turbulence energy dissipation rates, hence mixing rates, were inferred from experimental data.  The new code makes use of a k-ε CFD simulation of the flows and rates of turbulence energy dissipation and a previously published relationship between turbulence energy dissipation rate and random C-D rate.  Since random C-D depends on representing the fluid by discrete fluid elements that flow and mix, mass rates of their movements from segment to segment must be equal to the simulated flow rates in order to maintain material balance. 

                The C-D method for simulating effects of mixing on multiple chemical reactions has the advantage that no closure hypothesis need be used.  If the number of fluid elements per segment is high enough to be a statistically significant sample, then each element may be treated as a small batch reactor between each mixing event.  C-D mixing events must be frequent enough to adequately simulate the progress of the mixing and chemical reactions.  The method of simulation is tested by comparison of its results with experimental data.

                Advantages and disadvantages of the C-D method as compared to other methods, such as complex reaction closures, large-eddy modeling, direct numerical simulation and lattice-Boltzman simulation, will be discussed.  Example simulations using C-D for tubular and stirred-vessel reactions will be used to show how the method works.