Process of Forming Seamless Capsules by Concentric Nozzle System t

Adaptation of the natural phenomenon of drop formation has resulted in the de velopment of forming seamless capsules. The encapsulation behavior of gelatine in coolant fluid has been studied using a concentric nozzle. The major assumption was that capsules were formed through a wave-like instability in the concentric streams of extruded core, shell, and coolant fluids. Predictions of flow rates and fluid properties were compared with experimental results and were shown to be in reasonably good agreement. It was found that for a given set of flow rates and fluid properties optimum conditions could result in form ing uniform size capsules. experimental findings on this influence with the acid of a fundamental model concept.


Introduction
One of the advantages of a capsule is to apparently treat formless liquid as a solid form. A soft capsule which is formed by concentric nozzle system based on an orifice method is called a seamless mini-capsule, and has been widely used for a variety of industries such as food, pharmacy, and so on. This capsule shows excellent powder characteristics of fluidity and packing behavior since it is formed to have spherical shape. However, the influence of physical property on capsule formation process has been scarcely investigated. This paper presents experimental findings on this influence with the acid of a fundamental model concept.

Experimental equipment
The experimental equipment used in this work is schematically illustrated in Fig. 1.
Shell liquid and core liquid stored in each t This report was originally printed in J. Soc. Powder Tech., Japan, 20, 723-727 (1983) in Japanese, before being translated into English with the permission of the editorial committee of the Soc. Powder Tech., Japan.
32 tank CD and ® respectively were introduced quantitatively into a concentric nozzle ® by gear pumps @ and @ and discharged out of the tip of the nozzle so that a jet flow @ could be formed in a guide tube (]). When the flow of coolant oil filled in this guide tube was laminar, the stable elongation of the jet flow would be permitted to make sequential formation of resultant capsules. In this process, moreover, a vibration ring @ was equipped to form uniform size capsules. On the other hand, the coolant oil to be used for forming capsules by cooling was introduced into the guide tube (j) mounted at the center of a cooling cylinder @, after being cooled in a heat exchanger @. This liquid accompanied by formed capsules was transported through a cooling tube @ , before being made to be separated from the capsules by separater @ placed at the outlet of the tube. After this separation, the liquid was stored in a coolant oil tank (jJl and returned to the heat exchanger through a pipe @ by gear pump @.

Experimental procedure and results
In this work, the influence of each liquid flow rate and shell liquid viscosity on capsule formation process was experimentally investigated. For simplification, the core liquid, the shell liquid, and the coolant oil will be called  liquids 1, 2, and 3, respectively, in this paper. The formation of capsules is schematically outlined in Fig. 2. The jet length L is considered to be a distance where the jet flow of liquids 1 and 2 is capable of maintaining its stability. In this case, the diameter of this jet flow at the point which is L/2 distant from the tip of the nozzle may be reasonably regarded as a standard dimension of the flow; however it should be noticed that the diameter at an arbitrary point is dependent upon the vertical distance in the flow.

Relationship between jet diameter ~ and capsule diameter de
The influence of jet diameter d; on capsule diameter de was experimentally studied under the conditions listed in Table 1. As shown in Fig. 3, plots of the results hardly deviated from the single straight line which demonstrated that de would be twice as large as Cf;. It is found that this result might be in good agreement with the solution derived from the Rayleigh theory for instability of liquid jetsl).  Consider a typical capsule model during its successive formation as illustrated in Fig. 4. Provided that the distance between the successively formed two capsules (pitch P) is assumed to be the length of the cylindrical jet flow required to make one capsule, the following equation is given: As shown in Fig. 5, the experimental results based on the conditions presented in Table 1 seem to be little apart from the linear relation of Eq. (1 ). It follows that Eq. (1) could reasonably relate the jet diameter ~ and the capsule diameter de, which was also confirmed by Dabora 2 ). From Eq. (1 ), the capsule diameter de can be rewritten as a function of pitch Pas well as jet diameter de as follows.
This equation states that capsule diameter de can vary with jet diameter~ alone, if pitch Pis held constant. Pitch P may be held constant, if vibration is applied to the jet flow. By the way, under the conditions listed in Table 1 the encapsulation processes observed is schematically illustrated in Fig. 6; the capsule diameter decreased with increasing the value of q 3 . On the other hand, it was found that the jet diameter di increased with increasing the values of q 1 and/or q2.
Relationship between flow rate ratio ( q 1 + q2 )/q3 and jet diameter~ From the facts mentioned above, it can be known that liquids 1 and 2 might be positive factors for rising the jet diameter ~ and liquid 3 might be a negative one. Figure 7 shows the q,,=16.1 g/s q,=26.3g/s q,=35.0g/s q,=44.9g/s q,,=544g/s   dependence of the flow rate ratio (qt + q 2 )/q 3 upon the jet diameter d;.
When it is assumed that all the liquids are Newtonian and flow at the same rate, the following equation can be given: where 'Yi is the specific weight in g/cm 3 for the i-th liquid. The maximum flow rate of liquid 3 is expressed as The experimental condition employed in this work was that 'Yt = ' Y 2 = 1 x I0-3 g/mm 3 , ' Y 3 = 0.94 X 1 o-3 g/mm 3 ' and d3 = 20 mm. Thus Eqs. (3) and (4)  Relationship between flow rate ratio ( q 1 + qz)/q3 and capsule diameter de To find the dependence of the flow rate ratio (q 1 + q 2 )/q3 upon the capsule diameter de, another encapsulation experiment was carried out under the condition of the temperatures of 22, 80, and 10°C for liquids 1, 2, and 3 respectively and the concentration of 25% for liquid 2. The experimental condition and the result are indicated in Table 2 with the schematic illustration of Fig. 8. It is found from this figure that the flow rate ratio (q 1 + q 2 )jq3 was approximately proportional to the capsule diameter de with a region of 0.02 < (q1 + q2)jq3 < 0.06. This implies that there could be an optimum flow rate ratio for formation of the capsule with a uniform diameter.
Influence of flow rate ratio q 1 /q 2 on mean capsule diameter de As sEown in Fig. 9, the mean capsule diameter de was almost independent of the ratio of the flow rate of liquid 1 to that of liquid 2, q1 /q2.
Influence of liquid 2 concentration on mean capsule diameter The mean capsule diameters are plotted against the concentration of liquid 2 as a function of (q1 + q2 )jq3 in Fig. 10. It is found that the diameter decreased with increasing the concentration and this tendency became more remarkable as (q 1 + q 2 )/q 3 increased. This is probably because the spherical formation of the capsules was enhanced both by the increase in the surface tension of liquid 2 due to the rise of its concentration and by increase in the flow rate of liquid 3 which resulted in reduced jet diameter.
The physical factors investigated in this work were the flow rate of the liquids and the viscosity of liquid 2 and they might be also dependent on a more fundamental factor, i.e. temperature. Therefore further study on the influence of temperature should be required to control the capsule formation process more precisely, although only the restricted range of the temperature effect was investigated in this work.

Conclusion
It has been little known how the two kinds of liquid are discharged out of the concentric nozzle into another liquid in a jet flow. In this work, we investigated the influence of these three liquids on the diameter of capsule formed in this solidification process in order to elucidate such a complicated fluid dynamic behavior. The results of experiments and observations are summarized below.
1) The capsule diameter de is about twice as large as the jet diameter~-2) The capsule diameter de is dependent upon the pitch P as well as the jet diameter ~' and thus de may be held constant if vibration is applied on the flow to fix the value of P. 3) There is an optimum flow rate ratio (q 1 + q2)jq3 which yields the capsules with a uniform diameter. 4) The capsule diameter de decreases with an increase in the viscosity of the shell liquid. 5) The diameter of stabilized capsules may be determined when the velocity of the jet flow coincides with that of the coolant oil. Based on this work, we intend to investigate the influence of other physical factors such as temperature, specific weight, and so on.  Instr., 38, 502 (1967).