Considerations
The number of bacterial particles ν collected by the fibrous air filter was tentatively expressed by Eq. (1) on the following assumptions.
1) The filter of glass fiber is composed of a number of sheets amounting to ξ per unit depth.
2) The particle has a probability of collision p with each sheet in passing the filter.
3) The particles, the number of the collisions of which statistically proved not to exceed m times, have a possibility of escaping out of the filter.
Rearrangement of Eq. (1) with some approximations resulted in a general form of Eq. (3).
Experiments
The experimental apparatus and procedure in which radio-isotope ³²P was used were similar to those reported previously.²⁾ In the present paper, however, effects of fiber diameter and volume fraction on the longitudinal distributions were mainly discussed. Furthermore, the experimental results were analysed by using Eq. (3)', which was derived from Eq. (3). It was postulated that the intensity n of β-rays was proportional to the number of bacterial particles.
The following is an example of the analysis:
From a straight line assumed to run through points A and B in Fig. 1, k was determined, and from the intersection of this line with the ordinate, n₀'. Using n₀' and k together with the experimental data of dn/dL relating the data points (C, D and E), m was determined from Eq. (3)'. The linear relation between m and L is apparent as in Fig. 2.
Effects of fiber diameter and volume fraction on the distributions are shown in Figs. 3 and 4. In these figures, the relations between m and L pertaining to each experimental condition are also shown. The same basis as given above was employed in drawing a line through each set of distribution data points.
The interval of time employed in Eq. (4), in which bacterial particles are expected to escape out of the filter successively, was computed from each probability function of Eq. (6). Eq. (4) shows a statistical mean time interval so long as each probability function remains unaltered. The interval T is related with the collection efficiency as shown in Eq. (5). In this calculation, the linear relations between m and L were assumed to hold beyond the experimental range of L, together with some approximations as are expressed in Eq. (13).
Figs. 5 and 6 show the results of the calculations. The limits of significance of these computations are indicated by broken lines. The curve for control in Fig. 5 was derived, assuming a long-penetration relation in the case of d_f=8μ.
Although the curves in Figs. 5 and 6 are considered to be of use as a basis on which to discuss the performances of fibrous air sterilization filter, they remain to be checked further by some appropriate experiments.

View full abstract

Isobaric vapor-liquid equilibrium data on the binary systems carbon tetrachloride-ethanol and carbon tetrachloride-n-butanol were determined at 760mmHg total pressure, using a modified Othmer still. The data thus observed and those in literature on the systems carbon tetrachloride-methanol and carbon tetrachloride-n-propanol were studied thermodynamically by means of the Herrington method. The results showed their consistency and correctness.
In order to predict the ternary vapor-liquid equilibria, involving any of these binary systems, these data might be presented by the Lu's empirical formula with three constants.

View full abstract