Page 300 - Packed bed columns for absorption, desorption, rectification and direct heat transfer

286

The mean error of the Eq. (195) is 1.6%.
The comparison of Eq. (194) and (195) shows considerable difference
in their coefficients as well as in their exponents. The packing material itself
does not influence the mass transfer coefficient, since it is either fully wetted, or
covered with naphthalene. The only significant difference between the two
groups of packings is their considerably different free cross-sectional areas
(equal to the void fraction). A brief examination of some of Zhavoronkov's data
shows that the exponent of M4 fells as the free cross-sectional area rises. This
is consistent with the much lower value of this exponent in Eq. (195) (0.16 as
opposed to 0.47), reflecting the large free cross-sectional area obtained with
honeycomb thin-wall packings.
In physical terms the mass transfer differs in different parts of the
packing. This is connected not only with the formation and gradual thickening
of the diffusion boundary layer in the packing channels, which gives a fall in
mass transfer as the channel length increases, but also with the different
behaviour of the horizontal and vertical surfaces of the packing elements. As the
packing free area and M& ratio decrease, the influence of the horizontal
surfaces becomes more powerful. In other words, the influence of the group h/dt,
is not independent of the free cross-section area, as Eqs. (194) and (195) show.
Theoretically, the influence of RBG also varies with the free cross-
section area. This is because the onset of turbulence in the packing channels
depends not only on the flow velocity, the gas properties and the hydraulic
diameter, but also on the flow constriction at the channel inlet and the
corresponding expansion at the outlet.
Using all existing experimental data (Table 28) with and without the
correction of Zhavoronkov [168], the following equations are presented [167].

With Zhavoronkov's correction:

(O7SOJ2 m
Ks H- (197)

The mean error of equation (197) is 7%, and the maximum one is 30%, only for
one of 148 experimental points.

Without Zhavoronkov's correction:

(1M0Me) 9S
^ (198)