56 Summary

Modeling turbulent reactive flow processes is a complex and still developing subject. Before undertaking detailed reactive mixing modeling, it is essential to carry out the analysis of various relevant time (and length) scales of the system under consideration. Such an analysis will be useful to evaluate different modeling approaches and to select the best-suited approach. For most liquid phase reactive processes, where reactions are fast compared to turbulent mixing, phenomenological models such as multi-environment models integrated within a CFD framework look promising. For many gas phase reactive processes like combustion, a mixture fraction approach with presumed PDF may be more suitable. Computationally intensive approaches like LES and DNS are more suitable as learning tools. These simulations can be fruitfully employed to validate some of the important issues of the other computationally less expensive models. For most multiphase reactive processes, interphase mass transfer plays the central role. Turbulent mixing is rarely important in such cases, except when reactions occur in one phase and form immiscible products. For such cases, models to simulate homogeneous turbulent reactive mixing are applicable. Additional models to simulate nucleation or to simulate transfer of immiscible product into another phase may be included. It must be remembered that the quality of results of reactive mixing simulations will depend on the quality of several input data such as rate constants, mass transfer coefficients, interfacial area and so on. These quantities are seldom known accurately for industrial processes under relevant operating conditions. Sensitivity with respect to values of input parameters must be examined before using the simulations for reactor engineering. Approaches to developing tractable computational models for simulating complex industrial reactive flow processes and ways of direct or indirect validation are discussed with examples in Parts III and IV.

0 0

Post a comment