Mesurements of effective thermal conductivities of wetted packed beds were made by using an unsteady heating cell.
Packings used were of sand, lead shots or crushed glass. Either castor oil or water was used as a wetting liquid.
As the result, an experimental formula, Eq. (5), for effective thermal conductivities, was obtained.

View full abstract

For the fixed-bed reactors such as adiabatic multistage reactors, autothermic reactors and externally cooled reactors, the authors studied on the process optimum condition. which means the condition where the necessary amount of catalyst in VR/F is minimized when the conversion is specified.
Considering the reaction whose rate at a given composition of the reaction mixture have maximum at a temperature, as in the ammonia synthesis and the catalytic oxidation in sulphur dioxide, the authors extended Konoki's treatment for the adiabatic multistage reactor^{3, 4)} in whose operation maximum temperature is fixed, into the more general form, Eq. (18) through Eq. (21). These results were sucessfully applied to the sulphur dioxide convertor of adiabatic 3-stage.
For the autothermic process with two flow paths (Fig. 5), the authors using the simpler design equation (30) instead of Eq.(24), proposed a graphical method of solution based on Picard's method.¹⁰⁾ By this method we can easily determine optimum Γ and T₁₀ As an example the authors showed the result of trial and error for the determination of T and T₁₀ for the autothermic ammonia synthesis process. The basic equations (32) for the externally cooled reactor can be also reduced to the simpler form, Eq. (35), and so the same method above described will be employed.

View full abstract

This work was undertaken to clarify the mechanism of mass transfer into a liquid film flowing over the surface of a sphere and to see whether liquid is mixed completely or not at the points of juncture between the spheres connected in a vertical row. Experiments were performed on the absorption of pure carbon dioxide by water with and without a wetting agent, in columns containing 1, 3 and 5 spheres, respectively.
Experimental data for single sphere are shown in Figs. 3 and 4, and those for multiple sphere in Fig. 6, as the plot of the HL values against the Reynolds number on logarithmic coordinates. Results with single sphere were in good agreement with the theoretical equation based on the assumptions of unsteady-state diffusion and parabolic velocity distribution in the liquid film, as shown in Fig. 5. The addition of a wetting agent caused no change in the absorption rate in the case of single sphere, and the HL values for both the runs with and without agent were the same.
Fig. 7 shows the correlation among the results obtained with multiple sphere, when pure water was used as a solvent. Good agreement was obtained between the data and the theoretical equation, derived from the assumption that the mixing of the liquid was complete at the points of juncture between the spheres. Fig. 8 shows a similar plot of the data obtained by Yoshida and Koyanagi.⁷⁾
When a wetting agent was added to water, the values of HL were higher than those for pure water and increased with the number of spheres, as shown in Fig. 9. The data obtained by authors and those by Lynn, Straatemeier and Kramers⁵⁾ were in fairly good agreement with the theoretical equation based on the assumption that there was no mixing of the liquid as it flew from one sphere to the next.

View full abstract

Since the publication of Sherwood and Holloway's paper, ⁵⁾ it has been believed that the height of a packed column has no effect on the values of HL or kLa and the variation of these values with the packed height, observed in the experimental column, results from the end effects caused by the gas-liquid contact above and below the packing. On the other hand, van Krevelen and others⁷⁾ state that kL is proportional to the -1/3 power of the packed height and a recent study by Fujita and others1) shows that kLa varies with the packed height to the -0.19 power.
This work has been undertaken for the purpose of making sure whether packed height has any effects on the values of HL. Experiments on the absorption of pure carbon dioxide by water were performed in two columns, 6.6 and 12.5cm in inside diameters, respectively, packed with 15mm Raschig rings, with the packed height varying from 0.05 to 1.50m.
Observed values of N.T.U. are shown in Figs. 2 and 3, plotted on logarith mic scales against the liquid mass flow rate, with the packed height as a parameter. To obtain the true values of N.T.U. from these data, it is necessary to make correction so as to eliminate the end effects, for which a new method is proposed and employed. In Figs. 5 and 8, the N.T.U. values corrected by the above method are plotted on logarithmic coordinates against water rate. Fig. 5 obtained with the 6.6cm column shows the presence of three regions, A, B and C, so classified according to the values of liquid rate and packed height. The N.T.U. values in regions A, B and C are proportional to the 1.0, 0.92 and 0.56 power of the packed height, respectively, as shown in Fig. 6. The values of HL in region A, therefore, are independent of the packed height, but the HL values in regions B and C vary at the rate of the 0.08 and 0.44 power of the height, respectively, as seen in Fig. 7 and Eqs. 3 to 5. The N.T.U. values with the 12.5cm column, shown in Fig. 8 are proportional to the 1.0 power of the packed height.The HL values, therefore, are independent of the height of the packing, as shown in Fig. 9, and can be represented by the same equation obtained for region A with the 6.6cm column. The variation of HL values in regions B and C with that of the packed height may be due to the non-uniformity of liquid distribution in the column.

View full abstract

According to Prandtl's suggestion, the author calculated the velocity distribution and the pressure gradient of the non-Newtonian laminar flow in the region near the entrance of a circular pipe. The results are summarized as follows:
1) Though the inlet length for the Newtonian flow, when n=1, was not in agreement with the results of the author's calculation, the change in the inlet length for the non-Newtonian flow, when the values of n were varied, could be roughly estimated as shown in Fig. 3.
2) The pressure loss in the Newtonian flow, when n=1, was in good agreement with the results of the author's calculation. Therefore, on the assumption that the pressure loss in the non-Newtonian flow with different values of n would be in good agreement with the experimental results, too, the correction of the pressure loss Δq for different values of n was made as shown in Fig. 5

View full abstract