(458h) Large Eddy Simulation of a Precipitate Flow With Qmom

A precipitation reaction with four competing solid phases was studied using our LES code in two different geometries.  One simulation uses intersecting pipes flowing across a static mixer, while the other uses a confined impinging jet reactor (CIJR).  The Reynolds number in the simulations is approximately 27,000 based on the inlet conditions.  The complex geometries contain short inlet sections, which does not allow for the CFD code to develop the expected turbulence. In order to develop this turbulence, a digital filter method has been used to generate artificial turbulence which accurately matches prescribed first and second order fluid statistics. To account for the subfilter mixing scale, the scalar variance and scalar dissipation rate are calculated using a two equation model, which transports the second moment of the mixture fraction and the squared mixture fraction gradient.  

The properties of the solid precipitate phase are calculated with the general population balance equation (GPBE).  To obtain numerical closure to the problem the quadrature method of moments (QMOM) is utilized.  The GPBE contains a full set of particle source terms including birth, growth & aggregation kernels, which are highly dependent on the local chemistry of the fluid phase.  The physics for this problem have a timescale separation, such that the aqueous phase ions are considered to be in equilibrium. These equilibrium concentrations have been pre-tabulated with an external software package to reduce computational cost. The GPBE also utilizes a subgrid multi-environment approach consisting of three delta functions, where the mixing of these environments is determined by the local scalar variance value.