In order to find the most suitable operating condition for concentration of sea water by means of the direct contact freezing process, a test plant that could be easily operated in a steady state was designed and built.
The correlation between linear growth rate of ice crystals and operating condition of this plant was derived from the analysis of experimental data as follows:
W_I=3βIDIf (α/TDI)
where, W_I is crytalisation rate (or production rate), I is ice content in brine, T/α is holding time of ice crystals, D_I is linear growth rate of ice crystals, and α, β are correction factors.
D_I was estimated by analysing the size distribution of ice crystals produced in the freezer during operation in its steady state.α was necessary because of the difference between the holding time of ice crystals and that of brine.β was necessary because the total surface area of the crystals varied with the change of the size distribution, even if the cumulative number of the crystals was constant.

抄録全体を表示

(1) In the schematic arrangement of the reversible rectifying column, it is desirable that it resembles the actual column as nearly as possible and yet be clearly reversible.
(2) The deviation in the relationship between temperature and heat input or output of an irreversible column compared with that of a reversible column was examined. In the irreversible column m (<1) and K (positive integer) were induced as parameters of the shape of the relationship between temperature and heat input or output.
(3) Fig.8 illustrates graphically how the required work is increased from a reversible column to a vaporrecompression column. This suggests that the work of compression in a vapor-recompression column can be reduced by the following mean:
1) Reduction of the reflux ratio, 2) Use of a rectifying column equipped with a secondary still or a secondary condenser, or the use of successive rectifying columns, 3) Modification of the tower construction for larger heat transfer area, higher heat transfer coefficient and lower pressure drop through the column, 4) Recovery of mechanical energy from the excess vapor of the plant, 5) Improvement of the compressor efficiency.
(4) Table 2 shows numerically the difference in work balance between the steam heating process and the vapor recompressin process. According to Table 2, the reflux condenser loss is recovered by the use of the vapor-recompression method. Next, in the case of the vapor-recompression process of the adiabatic column, the tower loss becomes the largest loss, and the adoption of a column equipped with a secondary still is a method which will further reduce the required work by reducing the tower loss. In the case of the vaporrecompression column epuipped with a secondary still, the still loss becomes the largest loss. Consequently, Table 2 poses questions concerning the most effective type of tower construction which reduces the still loss and hence, reduces the required work.

抄録全体を表示

Previously the authors derived the theoretical equations (1) and (2) for the vapor-liquid equilibria in a binary system by introducing the factor a into the equations proposed by Guggenheim and McGlashan. It was also shown that the calculated activities (P/P⁰) and the experimental values are in good agreement if suitable values are given to the two parameters α and η in those equations.
A method of predicting the vapor-liquid equilibria in a binary system, which does not require the knowledge of binary systems, is developed here by employing Equations (1) and (2). Many parameters are determined from the experimental values obtained from various binary systems at constant temperature, and the values of w_AAB/zkT and w_ABB/zkT (generally by expressed as w_XXY/zkT) are evaluated from the parameters by employing Equations (6) and (7). The values of _wXXY/zkT are classified according to the type of the component X in the triplet XXY, and plotted against component Y as shown in Fig.1. Fig.2 is formed by collecting the straight lines of Fig.1. The parameters α and η can be predicted from Fig.2 by means of the method illustrated in Fig.3 and Equation (11).
As the parameters α and η have been predicted, activities (P/P⁰) can be obtained from Equations (1) and (2), and the vapor-liquid equilibria at constant temperature can be predicted. If the parameters α and η are assumed to be independent of temperature, the vapor-liquid equilibria at constant pressure can also be predicted by the trial and error method. The predicted results are compared with experimental values in Fig.5-8. Both are in fairly good agreement and it is seen that the vapor-liquid equilibria in a binary system can be predicted by the method given in this paper without recourse to experimental data.

抄録全体を表示

Heat transfer within soldered packed-bed in a flow system was assumed to have two processes, namely:
(1) heat transfer like a fin, and (2) heat transfer like an ordinary packed-bed.
Next, the differential equations about this cylindrical type of the bed were obtained as follows:
(1)(2)
Temperature distributions were measured in detail with 5 samples of heat transfer units having 11mm. dia. steel balls soldered at different heights within 5″ dia. brass tubes and Eqs.(1) and (2) were checked by graphical differentiation of these temperature distributions, and the values of h_pat were obtained. They gave almost the same values. Therefore, the assumption mentioned above on the mechanism of heat transfer in soldered packed-bed could probably be maintained.
The hp values obtained had good agreement with those in literature.

抄録全体を表示

The quantitative relation governing the blending time in jet mixing was investigated experimentally by using sodium chloride and sodium hydroxide as tracer. The experiments were carried out in two cylindrical vessels with diameter 0.4 and 1m., respectively. Diameter ratios of tank to nozzle were varied from 36 to 400. Most of liquids used in the tests were tap water and a few were Ucon (Polyalkylene glycol synthetic lubricant) aqueaous solution.
It was found that mixing time was independent of the angle of inclination of nozzle. Based on the assumption that mixing time is the function of the mean velocity of liquid across nozzle, the dimensions of nozzle and vessel and the physical properties of liquid, a dimensionless equation has been developed which expresses m ixing time number T as a function of N_Ren and scale ratios. At N_Ren ranging from 5×10³ to 1×10⁵, T became independent of N_Ren and the equation was found to be of the same type as that obtained by others for Larger tank blending. It also suggested that the blending of liquid with impeller could be explained with this equation.

抄録全体を表示