Page 37 - Coulson Richardson's Chemical Engineering Vol.6 Chemical Engineering Design 4th Edition

20
x 2 (or x 3 )
F 1 CHEMICAL ENGINEERING
F 3 (or F 2 )
Feed x 1 x 2
P 1
x 3
T 1
T
P
Select
F 2 (or F 3 )
Direction of calculation
(a)
x 2 (or x 3 )
F 1
Feed
x 3 x 1
F 2 P 1
F 3 T 1
T
P
x 2 (or x 3 ) Select
Direction of calculation
(b)
Figure 1.8. Information flow, binary flash distillation calculation (a) Information recycle (b) Information flow
reversal
effect reverses the flow of information through the problem and removes the recycle; this
is shown diagrammatically in Figure 1.8.

1.9.3. Information flow and the structure of design problems
It was shown in Section 1.9.2. by studying a relatively simple problem, that the way
in which the designer selects his design variables can determine whether the design
calculations will prove to be easy or difficult. Selection of one particular set of variables
can lead to a straightforward, step-by-step, procedure, whereas selection of another set
can force the need for simultaneous solution of some of the relationships; which often
requires an iterative procedure (cut-and-try method). How the choice of design variables,
inputs to the calculation procedure, affects the ease of solution for the general design
problem can be illustrated by studying the flow of information, using simple information
flow diagrams. The method used will be that given by Lee et al. (1966) who used a form
of directed graph; a biparte graph, see Berge (1962).
The general design problem can be represented in mathematical symbolism as a series
of equations:
f i v j D 0
where j D 1, 2, 3,..., N v ,
i D 1, 2, 3,..., N r
Consider the following set of such equations:
f 1 v 1 , v 2 D 0
f 2 v 1 , v 2 , v 3 , v 5 D 0