Basic Models of Computational Mass Transfer

kuotsong yu,xigang yuan
DOI: https://doi.org/10.1007/978-981-10-2498-6_1
2014-01-01
Abstract:The computational mass transfer (CMT) aims to find the concentration profile in a process equipment, which is the most important basis for evaluating the process efficiency, as well as, the effectiveness of an existing mass transfer equipment. This chapter is dedicated to the description of the fundamentals and the recently published models of CMT for obtaining simultaneously the concentration, velocity and temperature distributions. The challenge is the closure of the differential species conservation equation for the mass transfer in turbulent flow. Two models are presented. The first is a two-equation model termed as (c'(2)) over bar-epsilon(c') model, which is based on the Boussinesq postulate by introducing an isotropic turbulent mass transfer diffusivity. The other is the Reynolds mass flux model, in which the variable covariant term in the equation is modeled and computed directly, and so it is anisotropic and rigorous. Both methods are proved to be validated by comparing with experimental data.
What problem does this paper attempt to address?