Effective diffusivities of packed beds containing crushed calcite, beach sand and lead shots were obtained by examining diffusion of non-flowing benzene and watcr vapor through them.
To compare these diffusivities with electric conductivities, specific electric conductivities of similar beds were observed as function of fractional void, and the result wasreprcsented by experimental formula as follows.
The ratio of D_e/D_f obtained from this investigation was in agreement with that of k /k_f,
D : Effective diffusivity of packed bed
D_f: Diffusivity of fluid
k : Specific electric conductivity of packed bed
k_f: Specifice lectric conductivity of fluid
ε: Fractional void

抄録全体を表示

As the authors discussed in the previous paper²⁾, the sharpness of classification of hydraulic cyclones may be markedly improved by combining them in series as shown in Fig.1. In such a system, as we call the two-stage cyclone system, the overflow-side back pressure P_o of the first cyclone I is not always the same as the underfiow-side back pressure P_u, because pressure drops through the two second cyclones (ΔpII_o and ΔpII_u) may differ from each other. Consequently, flow rate and flow ratio through cyclone I will be functions of the overflow-side pressure drop Δp_o (=Feed pressure P_f-P_o), the underfiow-side pressure drop Δp_u (=P_f-P_u) and the dimension of the cyclone.
To see the effect of back pressures on the flow rate and the flow ratio of cyclone I, an experimental apparatus as shown in Fig.2 was constructed. The underfiow nozzle of the cyclone I was enclosed in a back pressure chamber A. Overflow and underfiow rates were adjusted by valves E and D and the back pressures were measured with manometers C and F. Thus full range of flow ratio from zero to 100% could be surveyed with this apparatus.
Typical examples of the results are shown in Figs.3 and 4, where the observed points are plotted on the Δp_o/γ vs ΔP_u/γ chart. Numbers above the points show the flow rate Q_f and numbers in bracket under the points represent the flow ratio R_f. It may be seen from these charts that the points of equal flow ratios lie on the straight lines (equi-flow-ratio lines) which cross the origin. All the observed points fall between those two limiting lines which represent zero and 100% flow ratio. Inclinations of equi-flow-ratio lines are correlated to D_e/D_u and R_f as shown in Fig.7 and equation (6). D_o and D_u seem to have equivalent effects on the flow rate and the flow ratio. Similarly the points of equal flow rate lie on the straight line which ties the same flow rate points on two limiting lines. The flow rate Q_f is represented in equation (8) as a function of Δp_o, Δp_u and the dimension of the cyclone.
Applying these results, dimension, pressure drop and Newton's classification efficiency are calculated for the purpose of designing two-stage, circulation type hydraulic cyclone system. Results of the calculation are shown in Fig.12, 14 and Table 3.

抄録全体を表示

The author has reported before the theoretical explanation of heat transfer in laminar flow of a pseudoplastic fluid, having either cooling or heating effect, without change in phase. By this method the author has tried to explain the heat transfer in laminar flow of Bingham fluid. The velocity distribution in Bingham fluid flowing through a tube are separable into two parts: one is a plug flow and the other is a non-plug flow.
Consequently the author proposes two differential equations for heat transfer, which can be proved to hold on the grounds that the bondary condition of temperature r=r_p, is variable and that the function is in y-axis direction.
The author has obtained the average bulk temperature of outlet by the application of Equation 1.44, and has found that the logarithmic mean of difference between the average bulk temperature of a fluid and the temperature of a wall is theoretically correct.
Hence from the theoretical relation,
when
The average bulk temperature when a=0.5 has been obtained by means of Equation 4·5, and Figure 5 shows Nusselt number plotted against Graetz number, revealing their theoretical relation.
The author has given, for example, a theoretical solution to the case in which the velocity distribution is a rodlike flow and the wall temperature changes linearly in y-axis direction. The results are shown in Equations 3·5, 3·7, and 3·8.

抄録全体を表示

The investigations into heat transfer in Bingham fluid have already been proposed by Orr, Bonilla and Salamone. But these were confined to heat transfer in turbulent flow. Because of high viscosity of Bingham fluid, most of the researches were restricted to the case of laminar flow. Hence the author has made experiments on the heat transfer in diatomite slurries flowing in a copper pipe, whose I. D. was 1.06cm, O. D. 1.3cm, heated length 2m, cooled length 0.5, 1 and 2m.
The author's theoretical solution regarding Bingham fluid has been used in the arrangement of data, and Oyama's result has been adapted to the Reynolds number in Bingham fluid. The following equation has been obtained for laminar flow:
where
the author recomends the following equation for Repl>7000:
but
for 1200<Re_pl<7000 as shown in Figure 7.

抄録全体を表示

The hydraulic cyclone, as a hydraulic classifer, has rather good classifying characteristics, but its sharpness of classification is not satisfactory enough when compared with that of the ideal classifer. In this paper, possibilities are discussed of improving the classification efficiency of hydraulic cyclones by combining them in series. The two-stage cyclone employed in the experiment is the combination of three cyclones in which the first stage cyclone (I) is directly connected with two second-stage cyclones, one, in the overflow side (II_o) and the other in the undcrflow side (II_u) as shown in Figure 2.
The overflow of II_o cyclone (OO) contains most of the fine particles and the underflow of II_o cyclone (UU) contains coarser particles. The underflow of II_o (OU) and the overflow of II_u cyclonc (UO) consist mainly of particles of intermediate sizes. When the feed material is to be classified into two products, namely fine and coarse products, both OU and UO must be returned to the feed tank and circulated as shown in Figure 3. We call such a system "two-stage cyclone of circulation type."
In the former report¹), the authors discussed the classification characteristics of one stage hydraulic cychones and represented they in the form of generalized fractional recovery curves. If the fractional recovery curves of all the coyponent cyclones areknown, the classification characteristics of a two-stage cyclone may be estimated from equation (2), where R_I, R_IIu, and R_IIo are the fractional recoveries of I, II_u and II_o cyclones respectively for a certain particle size. The overall fractional recovery (R_II) for the two-stage, circulation-type cyclone is defined as in equation (5), and the ratio of accumulation of particles by circulation (λ) as in equation (4). Examples of calculated values of R_II andλfor two stage cyclones are shown in Figure 8.
Newton's classification efficiencies are caiculated and tabulated in Table 1 which would be obtained when powdered materials, having Rosin-Lammler's type of particle distribution of equation (8), are classified at the cut size of 20 micron by a series of one-and two-stage cyclones, It is concluded from these results that the classifcation efficiency of a hydraulic cyclone may be markedly improved by these methods.

抄録全体を表示