Numerical solution, of complex flow models

CHAPTER 7 NUMERICAL SOLUTION OF COMPLEX FLOW MODELS... 

The flow becomes highly complex in a spiral-wound module containing a feed-side spacer screen. Numerical solutions of the governing equations incorporating most of these complexities have been/are being implemented (Wiley and Fletcher, 2003) using computational fluid dynamics models (see Schwinge et al. (2003) for the complex flow patterns in a spacer-filled channel). 

Analytical solutions also are possible when T is constant and m = 0, V2, or 2. More complex chemical rate equations will require numerical solutions. Such rate equations are apphed to the sizing of plug flow, CSTR, and dispersion reactor models by Ramachandran and Chaud-hari (Three-Pha.se Chemical Reactors, Gordon and Breach, 1983). 

This equation has been derived as a model amplitude equation in several contexts, from the flow of thin fluid films down an inclined plane to the development of instabilities on flame fronts and pattern formation in reaction-diffusion systems we will not discuss here the validity of the K-S as a model of the above physicochemical processes (see (5) and references therein). Extensive theoretical and numerical work on several versions of the K-S has been performed by many researchers (2). One of the main reasons is the rich patterns of dynamic behavior and transitions that this model exhibits even in one spatial dimension. This makes it a testing ground for methods and algorithms for the study and analysis of complex dynamics. Another reason is the recent theory of Inertial Manifolds, through which it can be shown that the K-S is strictly equivalent to a low dimensional dynamical system (a set of Ordinary Differentia Equations) (6). The dimension of this set of course varies as the parameter a varies. This implies that the various bifurcations of the solutions of the K-S as well as the chaotic dynamics associated with them can be predicted by low-dimensional sets of ODEs. It is interesting that the Inertial Manifold Theory provides an algorithmic approach for the construction of this set of ODEs. 

The convective diffusion equations presented above have been used to model tablet dissolution in flowing fluids and the penetration of targeted macro-molecular drugs into solid tumors [5], In comparison with the nonequilibrium thermodynamics approach described below, the convective diffusion equations have the advantage of theoretical rigor. However, their mathematical complexity dictates a numerical solution in all but the simplest cases. 

A successor to PESTANS has recently been developed which allows the user to vary transformation rate and with depth l.e.. It can describe nonhomogeneous (layered) systems (39,111). This successor actually consists of two models - one for transient water flow and one for solute transport. Consequently, much more Input data and CPU time are required to run this two-dimensional (vertical section), numerical solution. The model assumes Langmuir or Freundllch sorption and first-order kinetics referenced to liquid and/or solid phases, and has been evaluated with data from an aldlcarb-contamlnated site In Long Island. Additional verification Is In progress. Because of Its complexity, It would be more appropriate to use this model In a hl er level, rather than a screening level, of hazard assessment. 

The computer program for the material balance contains several parts. First, a description ofeach item of equipment in terms of the input and output flows and the stream conditions. Quite complicated mathematical models may be required in order to relate the input and output conditions (i.e. performance) of complex units. It is necessary to specify the order in which the equipment models will be solved, simple equipment such as mixers are dealt with initially. This is followed by the actual solution of the equations. The ordering may result in each equation having only one unknown and iteration becomes unnecessary. It may be necessary to solve sets of linear equations, or if the equations are non-linear a suitable algorithm applying some form of numerical iteration is required. 

---