This paper has been written to complement the paper by Cohen and Metzner (1981). These authors developed a parallel channel model of the packed column which they regarded as a porous medium divided into wall, transition and interior regions. Their model is comprehensive, but involves a substantial empirical content. In the present paper an alternative model is proposed. This model is simple and is somewhat idealized, but it has a rational basis, and it leads to results which highlight the twofold nature of the wall effect.
As Cohen and Metzner note in their literature review, many authors treating the wall effect have emphasized one of its two aspects at the expense of the other. One group of authors have noted that the porosity of a packed column is of necessity higher in the immediate vicinity of the confining wall, where the solid particles cannot pack so well, than in the interior of the column. This allows a freer movement of liquid near the wall, and thus leads to a larger mass flux, for a given imposed presstwe gradient, than would otherwise occur. A larger group of authors have dealt with the increased friction at the wall, which leads to a smaller mass flux, for a given pressure gradient, than would occur in the absence of the wall. The no-slip requirement at the wall has a restrictive effect on the flow.