CZ201215A3 - Device for refining molten glass by centrifuging - Google Patents

Abstract

The glass-clarifying apparatus by centrifugation, in which the separation of bubbles from the glass (7) by centrifugal force, comprises at least one fining centrifuge comprising a rotating body (1). The cylindrical rotating body (1) has an overall interior space V, which is partially filled with glass (7) intended to be clarified with a V content and is connected to the ambient atmosphere to escape bubbles from the glass (7). The total internal space V.sub.0.n has a mean radius (r.sub.2.sup.) between 0.05 and 1.0 m; and a total height h.sub.0.n in the range of 0.1 to 1.5 m. The proportion of V content of the rotating body (1) filled with glass (7) to be clarified is to the total internal content of the rotating body. of the body (1) in a ratio of V / V.sub.0. = 0.20 to 0.80. The number of revolutions of the rotating body (1) of the clarifier centrifuge during centrifugation in the range of 10 to 200 rad.sup.

Description

Technical field

The invention relates to an apparatus for clarifying glass by centrifugation, in which the bubbles are separated from the glass by centrifugal force. The apparatus comprises at least one fining centrifuge comprising a rotating body of rotary shape which is provided with a bottom at one end, optionally a lid at the other end, and is equipped with a glass inlet and outlet means. During centrifugation, the inner space of the fining centrifuge body is partially filled with 10 glass and connected to the surrounding atmosphere to escape bubbles from the glass.

BACKGROUND OF THE INVENTION

The process of removing bubbles in glass melting, called fining, is usually the slowest process 15 requiring high temperatures, long times and the use of often toxic or environmentally undesirable components of the glass melting composition. Even a small amount of bubbles in the glass is unacceptable in terms of its quality. Efforts to accelerate this process and achieve high glass quality, as well as to significantly reduce energy consumption, have led to the use of fining agents (most commonly used today are antimony trioxide combined with oxidant and sodium sulphate often combined with a reducing agent), and possibly other methods, to speed up the bubble removal process. In addition to the traditional chemical effect of fining agents, which release gases from the glass into the bubbles and thus increase the size of the bubbles, which as a consequence of buoyancy rise rapidly to the surface, melt saturation with rapidly diffusing gases such as helium is used. alone, or in combination with the force of gravity, accelerates the separation of bubbles from the melt. Another means is to use ultrasound to promote bubble growth and coalescence [1]. Another possibility is to adjust the shape and flow in the melting chamber so that the bubbles in the gravitational field have the shortest or easiest way to the surface [2-3] ,

In the recent past, attempts have been made to use centrifugal forces to separate bubbles from the enamel, analogous to centrifugation of emulsions and suspensions. However, this process is to be understood rather as centering, since bubbles as inclusions of very low density move in the centrifugal field towards the center of rotation. A rapid and perfect separation process was envisaged, but the results of both methods were often not perfect. The glass was usually free of larger bubbles, but often contained a significant amount of small bubbles

35. dust, whose removal was not successful even by changing the parameters of the spin process, most often by increasing the speed. When using high revolutions, the demanding requirements for the strength of the rotating device due to the high pressure on its casing came up.

The use of centrifugal force has been tried before in some companies, especially according to the patents filed by Owens Illinois, Inc. from Ohio, US. There are several 5 of their patents dealing with the refining of glass in a centrifugal field.

U.S. Patent 3,819,350 to Owens-Illinois, Inc., US, published June 25, 1974, discloses a method and apparatus for rapidly melting and refining glass. It is a modular system that aims to greatly accelerate the entire process of melting and homogenization and thereby speed up and reduce the cost of glass production.

10a The first part of the apparatus is a melting aggregate. Here the trunk is fed, heated and melted.

The resulting melt-batch mixture is agitated with a stirrer. Melted glass is formed containing molten grains of sand, cords and a large number of bubbles.

This melt is fed to the second part of the apparatus. Heating electrodes and mixing devices are located in the second unit. Here, the glass is exposed to higher temperature and shear 15 'due to the rotating cylinder inside the aggregate. Quartz grains, cords, and all reactions will be dissolved. However, a large amount of gaseous inclusions remains in the melt.

This foamed glass is fed to the third part of the apparatus. Here the glass is exposed to centrifugal force. Bubbles are removed and the clarified and melted glass is discharged from the processing plant.

20. By using this apparatus, it is possible to achieve the same tons / day power as in traditional melting plants, but with considerable savings in the space required for the melter and the time required to melt glass of acceptable quality.

A detailed description of the third unit of this centrifuge apparatus is disclosed in GB 1 360 916, owned by Owens-Illinois, Inc., US, published June 24, 1974. Pre-melted glass with a large number of bubbles is poured into the centrifuge cylinder. A plate with several holes is placed on the glass inlet at the top of the cylinder. It is intended to distribute the flowing glass to the walls of the centrifuge cylinder. On the underside of the cylinder is located another similar ring, but with differently spaced holes - as usual - around the perimeter. It also has the task of drawing the glass back to the walls of the centrifuge cylinder

In a preferred embodiment, the distance between the top of the paraboloid and the bottom of the cylinder is at least half, preferably equal to the length of the paraboloid.

As in the previous patent, GB 1 416 027 of Owens-Illinois, Inc., US, published December 3, 1975. 6.1974, discloses a method and apparatus for refining molten glass.

The device is similar to the previous example: centrifuge cylinder with inlet and top 35 lid with holes. The outflow of glass from the centrifuge is different. In this case, the glass outlet is not in the center of the cylinder, but along its periphery through the outlet channels.

y

Several modes of operation of this device have been tested:

- the thickness of the glass layer is 1 inch (about 25 mm);

the thickness of the glass layer varies from 0.001 to 1 inch;

a constant thickness of the glass layer along the entire wall of the centrifuge;

5, - the glass is with or without the presence of the fins.

All of the above-mentioned processes and equipment are designed for continuous operation and high power melting units. While these methods and devices allow the glass to clear and remove large bubbles, they do not clear the glass of the dust bubbles, so that the glass is finely imperfect.

SUMMARY OF THE INVENTION

These disadvantages are overcome or substantially reduced by the glass clarifying device by centrifugation in which bubbles are separated from the glass by centrifugal force.

15, The centrifugal clarification apparatus of the present invention comprises at least one

clarifying centrifuge. It is an object of the present invention that the at least one fining centrifuge comprises a rotating body of cylindrical shape with a peripheral sheath, a bottom and optionally a lid having axial symmetry in the axis of rotation of the rotating body. The rotating body has a total internal space _of V, which are partially filled with glass content of about 20 to Kurčenou, refining and which is connected with the surrounding atmosphere for the escape of bubbles from the melt. The total interior space V _o has a mean radius r _o of 0.05 -1.0 m; and the total height h ₀ in the range

0.1 ~ 1.5 m. The content proportion of the rotating body filled with glass intended for refining is internal to the total content of the rotating body in the ratio V / V _o = 0.20 0.80. The rotational speed of the rotating body of the fining centrifuge during centrifugation is in the range of 10 ^ 200 rad.s ¹ .

2f>. The main advantage of the present invention is that when clarifying the glass by centrifugation, bubbles are separated without dissolving the bubbles in the glass. With optimum guiding of the glass melting by centrifuging, with the claimed radius and height range of the rotating body of the fining centrifuge, its defined filling with glass and with the claimed rotational speed of the body, spinning 30 'bubbles from the glass can be surprisingly short. thousands of seconds even for the smallest bubbles with a radius of 5x10 ' ⁵ m to 1x10' ⁴ m. This effectively removes even the smallest practically expected bubbles from glass with a radius of 5x10 ' ⁵ m to 1x10' ⁴ m, which has not been achieved.

The solution allows the use of more clarifying centrifuges. For example, molten glass

35. Continuously, it can be processed in parallel in multiple centrifuges with a time-shifted mode for the necessary time before moving on to the next step. The finings can be carried out in a very wide temperature range of 1000 ° C - 1600 ° C, with most of them

X industrial glass is a real range of 1200rd - .1500 ° C. Higher temperatures are designed for harder glass types, such as borosilicate glass or LCD screen glass. Lower temperatures are suitable for soft glasses, eg highly alkaline glasses or highly lead glasses.

Defined ranges of device size, glass temperatures and average bubble velocities were obtained by mathematical modeling and experimental tests.

The similarities between the behavior of bubbles in different glasses in terms of their separation from the melt, as determined by a series of independent measurements, are used for the procedure of simple transfer of optimal conditions from model glass, where basic knowledge of bubbles in rotating melt has been obtained. .

This transfer is accomplished by a set of computational data on a model glass, grouped into several empirical equations, and using temperature dependence of bubble rate, under conditions without application of centrifugal force, which is measured laboratory on the glass to be used for the centrifugation process.

The data obtained include the optimum rotational speed of a given rotating cylinder at a given melt feed and at a specified process temperature and the expected removal time of the estimated smallest bubble present starting from the known most disadvantageous position in the melt cylinder.

The original in this set of knowledge is also the transfer of optimal conditions to another type of glass. This transmission is based on;

2tf. (a) Determine the rate of bubble growth in the glass to be centrifuged in the laboratory under conditions without the use of centrifugal force.

b) Transmission of laboratory measurement results using empirical equations giving optimum conditions for model glass within a defined range of investigated conditions, i.e., critical bubble growth rate, for conventional glasses and 25 corresponding temperature range for refining temperatures, and for rotating cylindrical bodies of claimed dimensions and claimed filling.

A rotating body having a rotational shape with axial symmetry is, for example, a cylinder or a cone. The roller is the most advantageous design in terms of production and function in operation. The cylindrical centrifuge body is advantageous in terms of spin rotation alone, when the viscous liquid during centrifugation at high speed can form an almost uniform layer on the inner wall of the centrifuge body shell, contributing to uniform centrifugation and bubble separation at this glass thickness. The centrifuge body can rotate about its vertical axis, it can also rotate in an axis perpendicular to the vertical axis, and can also rotate at an inclined axis, with gravity being used to fill and discharge the body and rotate the rotating centrifuge body in this sense. Thus, the fining centrifuge has an axis of rotation that can be tilted from a vertical position to a horizontal position as well as an inclined position, but a vertical position is preferred.

4 '

This optimal and even adjustable claimed value range according to individual cases, which must be coordinated with each other, is the result of several years of research and development.

When the lower limit of the claimed mean radius r ₀ of the internal space of the rotating body 5 'is a clearing device occupy a smaller space.

At higher claimed mean radius r ₀ of the internal space of the rotating body, it can be assumed that although the device is robust and heavy, but will be possible with the same implementation fining glass centrifuge higher performance.

When claimed smaller height h ₀ of the rotating body, the rotating body may have a larger mean radius r ₀ 10 internal space, and a rotating body will have a shape similar to the disc, in which case e.g. be preferred that the rotating body is equipped with a bottom and a lid.

At higher claimed height h ₀ of the rotating body, the rotating body may have a lower mean radius r ₀ of the internal space, and the refining centrifuge will have the shape of an elongated cylinder or a cone, and in such applications will need to have a clarifying centrifuge lid.

The claimed range of filling of the rotating body with glass is such that the glass does not fill the entire space inside the rotating body of the fining centrifuge, and the resulting free space has the shape of a rotating paraboloid whose height increases with rotation and eventually resembles a cylindrical interlayer of constant thickness. The rotation speed, and the amount of glass in the centrifuge after filling the body, are controlled so that the glass does not fill the entire space inside

20th centrifuge cylinders. The rotation causes a radially increasing pressure gradient which forces the bubbles to migrate radially to the enamel surface.

At a smaller proportion of the fill of the rotating body with the glass than the claimed lower limit, the centrifuge will show too little power per fill. At a higher proportion of filling of the rotating body with viscous glass, above the upper claimed limit, 25 small bubbles may dissolve at the casing of the cylindrical rotating body, which would considerably prolong their complete removal time.

At a lower number of rotations of the rotating body than the claimed lower limit, it has been found that the condition of effective and sufficient bubble separation at the bottom of the cylindrical rotating body may not be met. At a higher number of revolutions of the rotating body, above the upper limit of the claimed limit, the small bubbles in the cylinder housing again dissolve, which would again considerably prolong the removal time. The demands on the centrifuge material would also increase. The rotation speed should be chosen such that an optimum radial thickness of the glass is formed on the surface of the inner casing of the rotating body of the fining centrifuge. In practice for continuous fining, the uncleaned glass flows into the rotating body 35 at a rate such that the same amount of glass remains there and that the glass flows down from the body predominantly at the central axis.

This invention removes bubbles relatively quickly compared to traditional spinning. While maintaining optimum conditions, the glass should not contain any bubbles larger than 0.1 mm; thus, the glass should be free of bubbles, since smaller bubbles than 0.1 mm in diameter are not expected in the glass. The average residence time of the glass

5. in the centrifuge is about 15 minutes or less. The spin time depends mainly on the temperature in the glass melt - glass.

For some cases the device, in particular a height h _0, and low mean radius r ₀ of the rotating body clarifying centrifuge may be a rotating body at the inlet end toward the open glass. This open mouth serves as a glass inlet opening. This assumes lo. smaller amounts of centrifuged glass, such as luxurious or special types of glass. This solution is contemplated in the continuous flow of the finings.

In most preferred embodiments of the refining centrifuge, the rotating refining body of the enamel inlet-side comprises a lid with at least one reversible closable glass inlet opening and at the enamel-face side comprises a bottom with at least 15 one closable glass outlet opening. The bottom and lid fining centrifuge is the most advantageous application of the invention which ensures a safe fining process in a confined space. The openings in the lid and bottom ensure an undisturbed or adjustable glass inlet and outlet and gas evacuation. The glass inlet and / or outlet openings may be located on or off the axis of rotation of the fining centrifuge body.

2Ó The fining centrifuge may be unheated in case of excellent thermal insulation of the cylinder or in the case of rapid fining of a smaller amount of fined glass by centrifugation.

In most preferred embodiments of the fining centrifuge, and in practical use for most of the glass melts in the temperature range of 1000 - 1600 ° C, and also according to technical and design possibilities, the fining centrifuge is heated by gas or electrically or microwave heating.

For most particular solutions, it is preferable that the rotating body, the bottom, the side walls and, optionally, the lid of the fining centrifuge for centrifuging bubbles from the glass is made of refractory ceramic or refractory metal or alloy and is thermally insulated and has an outer metal sheath. This solution is possible for most common glass melted 30 in the temperature range of 1000m, 1600 ° C.

In continuous fining, a continuous flow of glass through the fining apparatus must be ensured, i.e. a rate of enamel inflow corresponding to the enamel flow rate must be ensured and adjusted.

The rotating body may comprise at least one reclosable opening for the inlet 35 of the viscous liquid, in particular glass, situated preferably in the reclosable lid. The lid is not necessarily required in the case of a narrow rotating centrifuge body, and then the entire area is determined for the viscous fluid inlet. However, recovery is expected in most cases

In a reclosable lid, with one orifice for a small amount of viscous liquid, or more orifices for larger diameter inlets, and a larger amount of centrifuged fluid.

The rotating body may also comprise at least one reversibly closable glass outlet opening, preferably located in the bottom of the rotating body. The glass effluent can be located anywhere in the rotating centrifuge body, but this location is easiest for both production and operation. Depending on the size of the rotating body of the centrifuge and the amount of viscous liquid to separate the bubbles therefrom, these openings may be more than one in the body. In the case of small-diameter rotating bodies, for example conical-type bodies tapering at the bottom, it can serve as an outlet opening 1d for the entire bottom.

The rotating body may or may not be heated. Heating does not have to be realized in the case of a smaller amount of centrifuged viscous liquid or a rapid clearing of bubbles, where, due to the short residence time of the glass in the rotating body, heating of the fining centrifuge is also not necessary.

15th rotating body may be heated with, in the case of centrifuging molten glass at high temperatures and greater quantities, e.g. gas or _t electrically or microwaves.

The rotating body intended to centrifuge the bubbles from the glass is preferably made of a refractory ceramic or a refractory metal or alloy. These resistant refractory materials are used because the temperature of most of the glass when centrifuging and separating bubbles in the 20th glass by refining is in the range of 1000 m600 ° C. It depends very much on the type of glass, whether the glass is of a softer type, eg lead and highly lead, which is expected to have a lower clarifying temperature. The temperature of most conventional glasses during clarification will be about 1400 ° C, in case of harder glasses, eg borosilicate glasses, higher temperatures are expected.

The glass clarifying device by centrifugation can be designed for continuous operation 25 for a smaller amount of centrifuged viscous liquid, in particular glass, for example glass melted by hand or luxury glass. Here it is assumed that one rotating centrifuge body will suffice for clarification. The apparatus may also operate for a larger amount of centrifuged glass, where it is envisaged to utilize multiple rotating centrifuge bodies arranged so that their modes are offset in time relative to each other, e.g. in a carousel arrangement.

The fining effect of the removed gaseous inclusions can be enhanced by optimizing the speed, temperature, residence time, or a combination of these conditions, or by changing the dimensions of the fining device.

The uncleaned glass is fed to a rotating body of the fining centrifuge which rotates about a central, substantially vertical axis. The glass flows down on the inner wall under the influence of gravitational force

3I of the rotating body. The rate of rotation, temperature and inflow rate are such that the glass forms a radial layer on the inner wall of the rotating body and allows the clarified glass to leave the body through a suitable outlet opening, preferably in the radial direction, after the rotation has ended.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is described in detail below with reference to exemplary embodiments and illustrated by schematic drawings, of which a rotating centrifuge body represents:

Fig. 1 is an axonometric view of the rotating body of the fining centrifuge, without glass, where the body is open at the top, the bottom has a closed bottom, Fig. 1a is an axonometric view of the rotating body of the fining centrifuge when filled with glass; and Fig. 1c is an axonometric view of the rotating body of the fining centrifuge at the glass outlet.

Fig. 2 is an axonometric view of the rotating body of the fining centrifuge, without glass, with the body open at the top, with an opening in the bottom; Fig. 2a an axonometric view of the rotating body of the fining centrifuge when filled with glass; Fig. 2c is an axonometric view of this rotating centrifuge body at the glass outlet;

Fig. 3 is an axonometric view of a rotating body of a clarifying centrifuge, without glass, where the body has a lid with an opening, its bottom is closed,

Fig. 3a is an axonometric view of the rotating clarifier centrifuge body when filled with glass, Fig. 3b is an axonometric view of the rotating centrifuge body during centrifugation, and Fig. 3c is an axonometric view of the rotating clarifier centrifuge body at the glass outlet;

Fig. 4a is an axonometric view of a rotating clarifier centrifuge body, without glass, wherein the body has a 2Š hole in the lid and bottom, Fig. 4a is an axonometric view of the rotating clarifier centrifuge body when filled with glass; 4c is an axonometric view of the rotating body of the fining centrifuge at the glass outlet.

In the following figures, various dependencies related to the removal of bubbles from the enamel during clarification by centrifugation are shown.

Giant. 5 shows the dependence of the time taken to remove the smallest bubble from the glass to rotate.

Giant. 6 shows the dependence of the clarification time on the rate of growth of bubbles from the glass under conditions without the application of centrifugal force.

X

- with i

l l

· 1 «·

Giant. 7 depicts the optimal time to remove a bubble of radius

5x10 ' ⁵ and 1x10' ⁴ m on the average thickness of the enamel layer at 1400 ° C.

FIG. 8 depicts the optimum time to remove a bubble of radius

5 x ^{10 -5} and ¹ x ^{10 -4} m at an average glass layer thickness at 1450 ° C.

Giant. 9 depicts the optimum time to remove a bubble of radius

5x10 ' ⁵ and 1x10' ⁴ m at the average thickness of the enamel layer at 1500 ° C.

Giant. 10 shows the dependence of the optimum angular velocity of rotation of the rotating body on the average thickness of the enamel layer for a bubble having a radius of 5x10 ^-5 m when half-filling the rotating body with glass.

Giant. 11 depicts the product of the angular velocity of rotation of the rotating body and the optimum bubble removal time with a radius of 5x10 ⁵ m to the degree of filling of the rotating body with glass.

Giant. 12 shows the average bubble growth rates in the glass without the effect of centrifugal force and, optimally, glass rotating in a cylinder of radius of 0.25 m and with half glass filling as a function of temperature.

DETAILED DESCRIPTION OF THE INVENTION

Example 1

Crushing centrifuge

The glass clarifying device 7 by centrifugation can be designed for continuous operation.

A substantial part of it, the clarifying centrifuge, can be heated by gas or electrically, or by microwave heating (not shown). The rotating body 1, the bottom 3, the casing 4 and optionally the lid 2 of the fining centrifuge for centrifuging bubbles from the glass 7 is usually made of a refractory ceramic or refractory metal or alloy, with an outer steel jacket and optionally thermal insulation under the steel jacket. The centrifugal clarification apparatus 7 35 may include various types of clarifying centrifuge body 7, some of which are listed below.

FIG. 1 shows an axonometric rotating centrifuge body 1 for a glass refining device 7 by centrifugation. In the rotating body 1, the bubbles are separated from the glass 7 by centrifugal force without dissolving the bubbles in the glass 7. The rotating body 1 has a cylindrical shell 4, the upper surface on the side facing the glass inlet 7,

And the opposite bottom surface formed by the bottom 3. The rotating body 1 of the fining centrifuge has a rotational shape with axial symmetry coinciding with its axis 8.

Figure 1 shows an empty rotating body and a clarifying centrifuge, without glass 7, which is open at the top, and closed at the bottom by a bottom 3 without the glass outlet 7. The body 1 thus has an upper surface on the side facing the glass inlet 7. and the open upper surface 10 serves as the inlet opening 5 of the glass 7 into the total interior space V _{of the} rotating body 7.

Fig. 1a shows the rotating body 1 just after filling with glass 7. In this Fig. 1a the arrow 9 indicates the direction of the glass inlet 7 into this type of rotating centrifugal body 7, the final amount of glass 7 when filling the rotating body 1 containing V, and 15, the final glass level 7 in the body 1 to be fined.

In FIG. 1b, the arrow 11 indicates the rotation direction 11 of the fining centrifuge body 1 when the glass 7 is centrifuged, as well as the formation of a parabolic glass 7 on the inner wall of the housing 4 and the bottom 3 of the body 1 during centrifugation.

In Fig. 1c the direction 10 of the glass 7 outlet from this type of body 1 is indicated after the spin 20 has been completed and looking at the pouring out of the glass 7 through the upper open surface of the fining centrifuge body 1.

FIG. 2 shows another type of clarifying centrifuge body 1, where the body 1 is again open as above, so that the open top surface serves as an inlet opening 5 of the glass 7 into the interior space V _{of the} body 1. However, in this case the centrifuge body 25 1 has a bottom 3 provided with an outlet opening 6.

FIG. 3 shows another type of clarifying centrifuge body 1, without glass 7, wherein the body 1 has a closed bottom 3, without an outlet opening. However, the body 1 is provided on its upper surface with a lid 2 with an inlet opening 5 for glass inlet 7.

FIG. 4 shows another type of refining centrifuge body 1, without glass 7, whose bottom 30 is provided with a central outlet 6 for glass outlet 7, and the lid 2 is provided with a glass inlet 5 for glass inlet 7.

As in FIG. 1a, FIGS. 2a, 3a, 4a show the respective type of clarifying centrifuge body 7 described, each time it is filled with glass 7 which occupies the

In these Figures 1a, 2a, 3a, 4a, the arrow 9 indicates the inlet direction 9 of the molten glass 7 into this 35 'type of clarifying centrifuge body 1 through the respective inlet opening 5. The inlet opening 5 represents either the entire open top surface of the respective body type 1 as shown 1a, 2a, or the inlet opening 5 is formed as a central opening in the upper reciprocating body cover 2 as shown in FIGS. 3a, 4a. These figures 1a, 2a, 3a, and 4a also show the level of enamel 7 in the body 1 intended to clarify the enamel 7.

Similar to FIG. 1b, FIGS. 2b, 3b, 4b show the respective described type of fining centrifuge body 1 at the moment of rotation of the glass body 1 with the glass 7. The direction of rotation 11 'of the body 1 is indicated by an arrow. All types of bodies 1 rotate about the vertical central axis 8 of the body 1. It can also be seen from the figures that the glass 7 forms a paraboloid when the body 1 is rotated.

2c, 3c, 4c, the respective described type of clarifying centrifuge body 1 is shown, at the moment of pouring of the glass 7 from the outlet opening 6 of the body 1. The direction 10 of the glass outlet 7 from the body 1 is indicated. 1c, 2c, 3c, 4c with an arrow.

The glass 7 flows out of the body 1, which has a solid bottom 3 when the body 1 is tilted, either from the open top surface of the body made of the inlet opening 5 in the lid 2, which also become outlet openings 6 as shown. 1c, 3c.

In the case where the bottom 3 is provided in the bottom 3 with a central spout 6 for the spout, the glass 7 flows out of this spout 6 in the bottom 3 by its own weight, without tilting the body 15 1 as shown in Figs. 2c, 4c.

4, 4a, 4b, 4c with the inlet opening 5 in the lid 2 of the body 1 and the outlet opening 6 in the bottom 3 of the body 1, both in terms of the manufacture of the body 1 and its function.

It is useful to note that the most advantageous embodiment for the production and operation of these 20 types of bodies 1 is that the outflow opening 6 and / or the inlet opening 5 is situated directly in the axis 8 of the body 1, the axis 8 being simultaneously the axis of rotation of the body

Although not shown, the upper surface of the body 1 on the side facing the glass inlet 5 having the lid 2 may be provided with a plurality of closable glass inlet apertures 5, in the case of a wider type body 1 or body 1 of larger dimensions.

In similar cases, although not shown, the bottom 3 may also be provided with a plurality of outflow openings 6. It will be appreciated that in these examples the outflow openings 6 and / or the inlet openings 5 are situated outside the axis 8 of the body 1Na in Figs. 3, 4, the total interior space V _{o of the} various types of cylindrical body 1 is shown, and for this total interior space V _o the total height h ₀ 30 and the corresponding mean radius r ₀ correspond.

Figures 1a, 2a, 3a, 4a show the contents 1 / of glass 7 to be clarified when the body 1 is filled.

For all the mentioned types of clarifying centrifuge bodies 1, the total internal space V _{o of the} body 1 has a mean radius r _o in the range of 0.05 µm; and a total height h ₀ in the range of 35 0.1-d, 5 m.

i t «

The proportion of the content V of the inner space of the rotating body 1 filled with the glass 7 to be fined to the total inner content V _{o of the} rotating body 1 is in the ratio V ₇ / V _o = 0.20 h-0.80.

An important parameter for the centrifugation of the glass 7 by centrifugation is the speed of the rotating 5 ' of the clarifying centrifuge body 1 during centrifugation, which is in the range of 10 - 200 rad.s ^-1 .

The glass refining device 7 by centrifugation works as follows:

For multicomponent bubbles - and only such ones occur in glass production - two lú mechanisms are used to remove bubbles from the glass 7, but only one is acceptable to meet the requirement of rapid clarification: centrifugation. The area of permissible conditions for a rapid centrifugation process has been found, which, in addition to the process parameters itself, is also dependent on the type of glass or the transport of gases between bubbles and glass 7.

In particular, the occurrence of very small bubbles in the glass 7 after centrifugation greatly limited the applicability of the glass finings for demanding glass manufacturing. Rather, it was possible to use centrifugation where the requirements for glass quality are lower (the standard for some glasses allows very small bubbles) or to use centrifugation only as the first stage of the bubble removal process. However, both options are a concession from the possibility of using centrifugal force as a full-fledged tool to obtain a completely bubble-free glass. The requirement for a successful spinning operation is to obtain, in a period noticeably shorter than that of conventional fining in the gravitational field, the glass 7 completely free of bubbles, even without very small bubbles - puff. However, detailed knowledge of bubble behavior was missing for this purpose. Indeed, when applying the centrifugal force to remove the bubbles, ratios occur different from the behavior of droplets or solid particles which are normally separated by centrifugation. Due to the compressibility of the gases, the bubbles 25 contract at the moment of initiation of rotation and the effect of the centrifugal force is reduced as it is dependent on the square of the instantaneous dimension of the bubbles. Increasing the bubble pressure also results in an increase in the concentration of the individual gases present in the bubbles on the surface of the bubbles according to Henry's Law, thereby promoting the dissolution of the gases from the bubble into the glass 7. Dissolve 3Ů completely. If complete dissolution does not occur, the bubbles will again increase in proportion to the original pressure increase after the rotation has ended, which appears to be a defect. This phenomenon was the likely cause of the occurrence of a larger number of small bubbles after centrifugation, as reported by the aforementioned patent solutions. Thus, for the centrifugation of the bubbles, their compressibility and subsequent expected interaction with the glass 7 appears to be essential for the efficiency of the process. Also, the possible mechanism of complete bubble dissolution requires its evaluation.

12.

It is clear from the foregoing that a successful technical solution is not so much based primarily on the technical arrangement of the fining plant itself, for example whether it is intermittent or continuous operation, as on accurate and general knowledge of the behavior of bubbles in a centrifugal field. A recent study by the Applicants of this patent utilized a mathematical model of the behavior of multicomponent bubbles in model glass for television screens to parametrically study the behavior of bubbles in a centrifugal field to explain the supposed complex bubble behavior under these conditions and to find its general features. The use of model TV glass was mainly motivated by the fact that a large amount of data was known for this widely produced glass at that time, especially the data on concentrations, Id solubilities and diffusion coefficients of gases required to revitalize the bubble behavior model.

A comprehensive data system obtained by the systematic mathematical modeling of the behavior of critical size bubbles on a model glass is processed with the aim of transfer to another type of glass and presented in the form of simple semi-empirical and empirical equations. The principle of transfer of data 15 to other glasses is to utilize the proven similarity of bubble behavior in different types of glasses, if the rate of bubble growth in the glass 7 under conditions without application of centrifugal force (measurable by laboratory) is the same for model glass. Based on the measured average bubble growth rate or temperature dependence of the average bubble growth rate under the assumed conditions (temperatures and / or pressure) in the glass designated for application and the requirement for effective fining performance of the apparatus or design requirement for the rotating body 1 and the optimum angular speed of rotation of the cylinder and the appropriate optimum time to remove all bubbles, including small bubbles of specified critical size, are obtained from the model data system.

For the mathematical model, a rotating body 1 in the form of a cylinder shown in FIG

Fig. 1, 2, 3, 4. The results of monitoring the bubble behavior in the glass 7 in the cylinder have shown that partial filling of the cylinder with glass glass 7 must be used, so that bubble removal from glass 7 by centrifugal force - centering - takes place in the glass layer 7. in the case of the rotating cylinder housing, and that the spinning conditions must be set quite precisely. The reason for the precise adjustment was that during rotation, the bubbles were actually removed by two mechanisms: by completely dissolving the small bubbles in the glass 7 and centering the larger bubbles due to their movement toward the curved surface. because the dissolution times of the small bubbles were too high compared to the times normally required for centrifugation separation, the dissolution of the very small bubbles occurred at higher revolutions at the cylinder jacket and as the speed increased, 35 larger bubbles began to dissolve but needed more At very low speeds, on the other hand, the effect of the centrifugal force was small and the separation of bubbles by centering was also slow.

From this behavior it was clear that the dissolution of the small bubbles should be avoided while ensuring a sufficient centrifugal force effect so that the bubble separation was rapid. Figure 5 shows, by way of example, that under the given conditions, ie glass composition, temperature, pressure and initial bubble size, the time required to remove the bubble from the rotating glass melt layer 7 is a relatively complex function of centrifugal force, which 0.5 m half filled with the melt of television glass represented by the angular speed of rotation of the cylinder. The minimum time at relatively low revolutions, expressed here by the angular rate of rotation at rad ^-1 , shows the conditions for the fastest removal of the bubble by separation, by centrifugation to the glass level 7, while the maximum shows the dissolution of the bubble. At an even higher speed, the bubble then dissolves and the pressure on the housing 4 of the rotating body 1 increases. If a whole set of bubbles of different sizes is present, dissolution always occurs at the smallest bubbles first and therefore the optimum conditions given by the 1S optimum , associate with the existence of the smallest bubbles or maximum bubbles present, which still allow for the dimension of such smallest bubbles to have a radius of 5x10 ' ⁵ m, or

1x10 ' ⁴ m at lower glass quality standard. Based on laboratory experience and examination of bubble-like defects, bubbles in molten glasses usually do not occur at a given melting stage, due to their previous residence in glass at higher temperatures when bubbles very slowly grow. If we look at these smallest bubbles in the search for optimum conditions, we find the strictest conditions for the case.

In the curve of FIG. 5, the worst-possible state of bubble removal and the change of the spinning mechanism, i.e. the change from spinning to bubble dissolution, are shown at the highest point. The curve to the left of the highest point represents the state where the bubbles are in the glass 7 '25. centrifuged.

The curve to the left of FIG. 5 also shows the ideal state at the lowest point of the curve, corresponding to the shortest time of bubble removal by centrifugation. The curve to the right of the mechanism change, from the highest point of this curve, is the state in which the bubble dissolves in the glass 7.

The foregoing evidence suggests that the setting of suitable spinning conditions is limited to a relatively narrow range of magnitude of centrifugal force action. As already assumed, the effect of centrifugal pressure caused not only the shrinkage of the bubbles, which is unfavorable in terms of centrifugation, but also caused the diffusion of gases from the bubbles into the melt - glass 7, further reducing their size and the effect of centrifugal force. The second effect was then 35 'dependent on the number of gases present in the bubble and the rate at which the gases were able to diffuse through the molten glass 7. Thus, it would seem that the use of centrifugal force in such a situation is associated with too high requirements for its adjustment process. however, modeling results

1 <

studies have shown that, under favorable conditions, the rate of bubble separation process, as expressed by the fining times of the smallest bubbles, is considerably higher than when the bubbles are released by gravity only. Thus, for the practical application of the centrifugal force to remove bubbles from the molten glasses, it is necessary to obtain a generalization of the above-mentioned bubble modeling results so that an optimum centrifugation condition can be found for any glass 7 by the generalized results.

For practical use it is necessary for the type of glass correctly select any refining additives and further conditions of centrifugation: radius r _Q and a height h ₀ of the rotating total cylindrical inner space V _of the body 1, its degree of filling glass 7 and a rotational speed of Q1 inner space V _of . It is also necessary to estimate the critical highest fining times of the smallest bubbles.

The present invention relates to intermittent clarification, which is particularly important for glasses produced in smaller quantities or for use in continuous operation, where the removal of the clarified glass 7 can be achieved by using a plurality of rotating cylinders of the bodies 1, 15 optionally with the clarified glass container 7.

Setting appropriate spin conditions for each glass would require mathematical modeling of bubble behavior in the anticipated rotating space and finding optimal conditions by repeated calculations. However, the need for large amounts of data, in particular gas data dissolved in glass or bubbles, makes this possibility almost impossible, as the necessary data are obtained by special measurement methods and their acquisition requires long-term measurements. Experience from existing patented and known solutions then shows that the estimation of parameters without detailed knowledge of bubble behavior does not lead to a quality bubble removal process 7. However, the experience of the present inventors shows that bubbles behave in glasses very similarly during their removal from the 25th glass. 7, by sailing to the surface due to buoyancy. Many measurements have shown that even when using different fining additives on different types of glass and under different convenient temperature or pressure conditions, the bubbles have very similar fining times to rise to the surface if they exhibit the same rate of growth of their dimensions. Thus, under conditions of rational operation of the fining process, the bubble growth rate, which determines the instantaneous dimension of the small critical bubbles, critical to the time of their clarification and the other properties of the glass 7 important for bubble growth, such as its density and viscosity, play only a minor role. This also means that the rate of bubble rise in the different glass 7 or under different temperature conditions does not differ greatly if the bubble growth rates are identical, as determined experimentally. This is evidenced by the dependence of the fining time of very 35 small bubbles of initial radius and ₀ = 5x1 CT ⁵ m on their growth rate in Figure 6, obtained on different types of glass under different temperature and pressure conditions. In the wide range of bubble growth rates, the fining time is a function of their rate of growth, and thus the bubble growth rate is the only necessary and very reliable indication of the so-called clarity of the glass. Practical experience has shown that poorly cleared glasses exhibit bubble growth rates of less than 10 ' ⁷ m.s ^-1 , medium-clear glasses of between 10' ⁶ <10 ' ⁷ m.s' ¹ and very well-clear glasses X values less than over 10' ⁶ m ^{/ s} . The bubble growth rate is easily measurable in the laboratory by high-temperature monitoring of the bubble radius versus time and subtracting the slope of the near-linear bubble-time relationship. The average bubble growth rate is then obtained from the measurement of more bubbles. This fact of the similarity of the glasses in the fining step by gravity can also be used to find the optimum lo conditions by centrifuging them.

Since even in a centrifugal field, bubbles are removed by a mechanism based on the difference in bubble and glass density 7 (the radial bubble velocity equation is formally the same as the Stokes bubble equation for gravity field), it is very well assumed that bubbles with the same bubble velocity they will behave very similarly in a centrifugal field. This assumption is based on the results of mathematical modeling in a centrifugal field, which have shown that a very important parameter of bubble removal is primarily their ability to dissolve or grow, which is generally determined by supersaturation or saturation of the glass 7 under given conditions without applying centrifugal force. This ability is well represented by said bubble growth rate, which is laboratory measurable. Another important and general factor of the fining process already applied in the rotating cylindrical space is the centrifugal pressure generated by the rotation, which counteracts the supersaturation or degree of saturation of the molten glass 7 caused mainly by the application of the fining agent at the appropriate temperature. This centrifugal pressure may, under certain conditions, also cause partial dissolution of the bubbles or their complete dissolution. Centrifugal acceleration causes the bubbles to move radially toward the center 2 of the cylinder. The last important factor of the fining process in the centrifugal field then shows the time spent by the bubble in the molten glass 7 until it reaches a curved level. This time may be represented by the thickness of the enamel layer 7 which must be passed through the critical bubble starting from the cylinder housing. The role of individual factors is thus:

3Ó Bubble growth rate measured under given temperature and pressure conditions on a given type of glass without the effect of centrifugal force represents the general ability of bubbles to grow or

ii dissolve atmospheric pressure even under changed pressure conditions due to rotation.

The centrifugal pressure induced by rotation is a factor limiting the effect of supersaturation or saturation of gases, the value of which depends on the spinning parameters, i.e. the rotation speed, the 35th radius of the cylindrical space and the layer thickness of the glass during spinning. At the same time, centrifugal pressure provides information on the necessary strength of the rotating space, which must easily withstand the compressive load.

, «10« 4

The centripetal acceleration determines the speed at which the bubble moves towards the center of the cylindrical rotating body L

The thickness of the glass layer 7 of the casing 4 of the cylindrical rotating body 1 determines the bubble removal time.

The present invention is based on the assumption of similar bubble behavior in a centrifugal field, if the centrifugation conditions result in the same bubble growth rate values and relate to the same thickness of the centrifuged glass layer 7. If the conditions for the shortest critical bubble removal time can be defined. the smallest bubbles starting from the most disadvantageous position at the bottom of the cylinder 3 and at its jacket, centering towards IQ. to the center of the rotating cylinder that meet the minimum for practical use. These conditions are also transferable to other glass and other spinning process configurations. Similar bubble behavior then means that for critical bubbles of the same growth rate and the same average glass layer thickness, the critical bubble removal times in another glass will be approximately the same and also the optimum rotation rate of the cylinder will be the same.

15. In order to carry out such transfer, it was necessary to expand the number of modeled cases so as to cover a wide range of practically achieved bubble growth rates and glass thicknesses 7 in rotating cylinders. Different thicknesses of the glass layers 7 during centrifugation can be achieved either in cylinders of larger radius with smaller or average filling or in cylinders of smaller diameter and higher glass filling 7. The temperature range of 1350 1550 ° C, corresponding to a wide range bubble growth rate in the model glass at normal pressure 1,13x10 ' ⁷ ms' ¹ -4,62x10 ' ⁶ m.s' ¹ , radii of rotating bodies 1 of fining centrifuges from 0,1 to 0,75 m and the filling level of rotating body 1 in the shape of a cylinder by the ratio of the volume V of the glass 7 to the total internal volume of the rotating body 1 V7V ₀ from 0.25 to 0.75.

25, At each temperature, average bubble growth rates at normal pressure were subtracted from an area of approximately linear bubble radius versus time, representing both the growth rate and bubbles of Table 1 and available for laboratory glasses other than model glasses. 'i;

These values are shown in Table 1, together with the average critical bubble growth rate optimally obtained by modeling the centrifugation in a 0.25 m 3C radius cylinder (half-filled with melt of model TV glass) because it is important that bubble growth rates for non-centrifuged and centrifuged cases were similar, as shown in Table 1 below.

17.

Table 1

temperature [° C] growth rate [ms ¹ ) ^ 1350 1400 1450 1500 1550 ω = 99 _0p t [rad.s ^-1 ] critical bubbles during centrifugation 2,19.10 ^-8 3,39.10 ^-7 1,33.10 ^-6 2,71.10 ^-6 4,62.10 ^-6 ω = 0 [rad.s ^-1 ] critical bubbles without centrifugation 1,13.10 ^-7 4,34.10 ^-7 1,01.10 ^-6 2,08.10 ^-6 3,21.10 ^-6

Table 1 shows the average growth rate of bubbles of critical size 5.10 ^-5 m moving in a cylinder of 0.25 m radius, filling 0.5 under optimal conditions and at 5 different temperatures. Table 1 shows the average value of bubble growth rate in model TV glass when spinning glass and without spinning TV glass at ω = 0 rad.s ^-1

These results are also shown in Figure 12, where two dependencies are shown. One curve shows the average values of critical bubble growth rate in a model TV glass with an initial radius of 5.10 ^-5 m moving in a cylinder of radius 0.25 m with 0.5 0.5 Id filling, rotation under optimal conditions and at different temperatures. The second curve shows the average growth rates of the critical bubble in a model TV glass without rotation, at ω = Orad.s' ¹ , at different temperatures.

For temperature dependence of average bubble growth rate of model glass, as well as for easier expression of melt (TV glass), see Table 1, empirical equation 15 of eg polynomial or exponential type is created (Fig. 12):

da / άτ = Aexp ^ —j (1) where: a is radius [m] of bubble, b is constant [s], c is constant [s], τ is time [s], A is constant [ms ^-1 ] . The constants are obtained from experimental values.

By detailed modeling of individual cases, dependencies were obtained between times of clarification of critical bubbles of initial radius size and _o = ⁵ x 10 ^-5 or 1 x ^{10 -4} m and the angular velocity of rotation of the cylindrical fining space. From these dependencies, the minimum clarification time of the critical bubbles at the optimum angular rotation speed for practical use was subtracted (see minimum in Figure 5). The dependence of the shortest - optimal - time T _opt at a given temperature (bubble rate) is directly proportional to the thickness of the glass layer 7 in the cylinder over a wide

18.

range of cylinder radii and filling. This is demonstrated by Figures 7 * 9, for temperatures of 1400, 1450 and 1500 ° C, and for critical bubbles of radius 5x10 ^-5 and 1x10 ^-4 m; wherein the filling of the cylinder with the glass 7 is between 0.25 and 0.75.

For ease of expression, with respect to the following equations and calculations, the term melt is further used for molten glass 7, and for cylindrical rotating body 1, the term cylinder is further used wherever appropriate.

Fig. 79 shows the dependence of the optimal bubble removal time (bubble removal time at optimal angular rotation speed) on the average thickness of the glass melt layer in the centrifuge cylinder at temperatures of 1400, 1450 and 1500 ° C. The radii of critical bubbles are 5x10 ' ⁵ and 1x10' ⁴ m.

Thus, from the figures 7% 9 at a given temperature, it is possible to read the time of removal of the critical bubble at a given temperature and at a given thickness of the melt layer in the cylinder. Since, for a given melt, we find the laboratory bubble rate corresponding to any temperature within the temperature range investigated by modeling, it is necessary to obtain an empirical equation expressing the dependence of the linear dependence constants in Figures 7-9 on the model glass temperature.

The linear dependence of r _ppt on the average layer thickness is then:

2Ú τ _ορ , = k (t) d + q (t) (2) where: r _op is the optimum bubble removal time, k (t) is the temperature-dependent constant [s.m ' ¹ ], 3 is the average thickness [m] of the glass layer on the cylinder jacket, q (t) is the temperature-dependent constant [s].

Another necessary transfer value is the value of the optimum angular velocity ω _ορ ι, which determines the rotation speed of the cylinder in the applied case. An example of the modeled dependence of 2f (optimal angular velocity cú _opt on average thickness of the melt layer at different temperatures is modeling for a critical bubble with a radius of 5x10 ' ⁵ m, with a cylinder filling ratio V / V _o = 0.5 is shown in Figure 10. that at higher melt layer thicknesses in the cylinder, the value of ω _ορι depends very little on the temperature.

To obtain an accurate value of ω _ορΙ, the product values < _{jJ 0p} tT ₀ pt. Serve 30 which number represents the number of radians (and thus the number of revolutions) that the cylinder must perform in order for the bubble to reach the level. Modeling has shown that the values of this product for a bubble of given size, cylinder filling, and temperature are roughly independent of the radius of the cylinder and the condition can be characterized by the average value of the product. The values of the mean product ω _ορ , τ _{ορ (} for the critical bubble of radius 5x10 ' ⁵ m are then plotted as a function of the cylinder filling V7V ₀ in Fig. 11.

19Í

However, to obtain the corresponding value m _opt at any model temperature corresponding to the laboratory measured bubble growth rate, it is necessary to obtain the temperature dependence of both constants characterizing the linear dependence w _{op /} r _opf on the filling W _o in the form:

V ^a o _P , T _opl = k (t) - + q (t) (3) where: ω _{ορ (} is optimal angular velocity [rad.s ^-1 ] rotation of centrifuge cylinder, τορ! Is optimal time [s] removal melt bubbles [s], k (t) is the temperature-dependent constant [rad], V is the amount [m ³ ] of melt in the centrifuge cylinder, Vo is the volume [m ³ ] of the centrifuge cylinder, V / V _o is filling the centrifuge cylinder by melt, q (t) is a temperature-dependent constant [rad].

The corresponding value co _opt in the present case is then obtained from the value ω _ορι τ _ορ ι calculated by 1Ů from equation (3) with the already known value τ _ορι from equation (2). At the same time, the average centrifugal pressure at the cylinder casing corresponding to the minimum, ie the optimum angular speed of rotation, is calculated for the check according to the formula:

pX _P , X ~ øi)

P _a = --— z ---- ² (4) where: ρ _ω is the pressure [Pa] caused by the centrifugal force on the shell of the cylinder of radius r _D [m], ω _ορΙ is 1Š angular velocity [rad.s ^-1 ] rotation at optimum, p is the density [kg.rri ³ ] of the melt and ii »is the average radius [m] of the curved level.

It should be taken into account that the local pressure on the cylinder shell will be slightly greater than that calculated at the point where the critical bubble starts, ie at the bottom of the cylinder due to the curvature of the level. The pressure calculation is only performed to check that, under the optimum conditions, there is not too high a pressure on the cylinder housing (this pressure can be given by the prescribed value for the cylinder).

Thus, the specific procedure for determining the optimum spin conditions for a given case is as follows:

2 <

1. The laboratory measurement shows the temperature dependence of the average bubble growth rate at a given temperature or at several temperatures (temperature dependence). The temperature dependence of the average bubble growth rate, α, is expressed by an empirical equation, eg of the polynomial or exponential type, as in the following example melt model:

3Ó.

d [m / sec] = exp

58544 f + 273.15.

t 6 (1350; 1550) ° C (5) where: d is the average velocity [m.s ^-1 ], bubble growth t is the temperature [° C],

From the temperature dependence obtained, read the bubble growth rate at that temperature and subtract from the graph in Figure 12 or Equation (1) for this bubble growth rate.

20 May

an appropriate temperature that corresponds to the temperature of the melt at which the melt has the same bubble rate as the melt applied.

2. Select some spin parameters. Typically, the parameter selected is temperature

5) centrifugation resulting from the conditions of operation of the whole melting process and the radius of the rotating cylinder r ₀ , which is given by the technical possibilities of the designers and the arrangement of the whole equipment. Furthermore, it is necessary to determine the average thickness of the melt layer in the rotating cylinder, which would correspond to our performance requirements. For this purpose it is necessary to estimate the cylinder filling melt V / V _of the permitted range of 0.25 H.0,75. For the selected cylinder filling (it is recommended to use the first value of V / V _o = 0.5 as the vessel), the average thickness of the melt layer is calculated according to the equation:

(6) where: δ is the average thickness [m] of the glass layer on the cylinder shell, r _{0 the} radius [m] of the cylinder, W _o is the filling of the centrifuge cylinder with melt.

Subtract the third or calculate the value of y for a given value _OPŘ critical size if ₍₀ = 5x10 ⁵ m or 1x1 θ ^'4m) bubbles in the graphs in Figure 7-9 or from empirical equations derived by treating linear dependence r _O pt on the average layer thickness. These equations have the formal form of equation (2), where the parameters of the lines match the equations:

For a ₀ = 5x10 ' ⁵ m and guideline we have from model experiments:

(7)

_T is a constant [s.nf ¹ ], t is a temperature [° cj].

For y-segment:

q = 0; r e <1400; 1550>, <7 = 30 / -42000; / e <1350; 1400>

r _fí pl where: q _Topt is constant [s], t is temperature [° cj.

For a ₀ = 1x10 ' ⁴ m the slope k = 39,65 exp' tipl ¹ (9) where: fc _ropf is constant [s.m ' ¹ ], t is temperature [° c] (10) where: o, is constant [ s], t is the temperature [° C].

• O? and.

After obtaining the value r _opt, it is good to see the performance of the device by calculating the effective performance of the device, which is the expression:

^opl (11) where: Vgf is the effective power [m ³ .s ^-1 ] of the device, V is the volume [m ³ ] of the centrifuge cylinder, r _opt is X the optimum bubble removal time [s].

The value of V _ef [m ³ .s' ¹ ] after the values for V ar _opf are _reached depends on the cylinder filling approximately

í \ ^1/2 1 l · v \ 1- 1--

(12) where: Vgf is the effective power [m ³ .s ^-1 ] of the device, the I / _O is the melt filling of the centrifuge cylinder.

10- This function has a slightly decreasing character with increasing cylinder filling, the decrease in effective y power from filling 0.25 to filling 0.75 is 19, ^ / o, at filling 0.5 the drop is only 8.5%

Therefore, the recommended filling of the cylinder is around 0.5, while the smaller filling slightly increases the effective performance, but it will be necessary to empty the cylinder more often and fill with melt.

4. The w _opt value for the case is obtained. The situation is that the average product ω _ορί τ _{ορ (} at a given temperature and for a given critical bubble size increases within a given cylinder filling range of 0.25 ^ 0.75 linearly with its cylinder filling and the corresponding value is subtracted from Figure 11 (6) or equation (3) Empirical temperature dependence applies for both linear dependence constants within the given filling range:

For a ₀ = 5x10 ' ⁵ m we have the following model calculations:

^T i> pt

749.3 exp (500.6

J-1264,7 (13) where: Α · _ωορίΧο '. is constant [rad], t is temperature [° C]

For q from model calculations we have:

(14)

25. where: qw _0Ftt - ₀ .. _r is constant [rad], t is temperature [° C],

(15)

For a ₀ = 1x10 ⁴ m we get from model calculations:

k _T = 915expf ......- ⁴ - ^{, 6} Ί (J-1271.1J where: k. _r is constant [rad], t is temperature [° C].

Opopi ppl (410 5 Ί = exp 5.07 + --- t / -1275,4j (16) where: <7ω _{ορ (τβρί} is constant [rad], t is temperature [° C]

From the obtained values W _{0p 0} pt tr is then calculated by w _opt already known at t R _Op.

Thus, the conditions for centrifuging the melt are obtained. They are given by a set of values r _p , V / Vq,, hJ _O pt, T _O pt and V _e f.

10. This process will not usually give optimum conditions exactly, but will find an area of optimal conditions for the centrifugation process, which is not to be found by estimation only about the complexity of the bubble behavior in the glass melt and under the action of centrifugal force. It provides the conditions for the least favorable case in a given configuration and for a given spin parameter, thus working with some margin. Without this knowledge of behavior

15. Bubbles usually lead to only a worsening of the situation. Conversely, by monitoring the quality of the glass after the centrifuge, when applying the conditions obtained by this model, the conditions can be approximated to the optimum if the improvement requires:

I. If the glass behind the centrifuge is free of bubbles, there may be a reserve in the process and it is possible to carefully increase the speed of the roller or increase the filling of the roller.

2ό. II. If the centrifuged glass shows larger bubbles (tenths of a mm or more), the fining conditions are inadequate, the most suitable means is to reduce the filling of the cylinder with the melt, or it is possible to carefully increase the speed of the cylinder.

III. If the glass behind the centrifuge shows more very small bubbles, it goes very

25. Probably the case where very small bubbles in the cylinder shell dissolve, a suitable step is to reduce the rotation speeds or reduce the filling of the cylinder.

When clarifying on a laboratory scale, these imperfect conditions are reflected in the bubble distribution in the melt after cooling. The defects according to point 2 are manifested in the whole meat 3 (melt, possibly at the curved level, the defects according to point 3 mainly in the cylinder jacket).

When applying the results, the issue of efficiency of centrifugation under given conditions has to be addressed. It is possible that the conditions for the application of the spin will give the critical bubble removal times higher or comparable to those in which

23I, rotation applied. Therefore, the obtained value T _opt for the critical bubble must always be compared with the value of the removal time of the same bubble under the condition of free ripple in the gravitational field (the roll does not rotate and the melt forms a static layer in the roll). The bubble removal time of the initial radius a ₀ , the growth rate α from the static melt layer of thickness / outlet d from the bottom is given by the equation:

¹⁷ \ ^{1 =} Xý ( ^σ ο ^{τ + β} ° ^{άτ2 +} x ~) ( ¹⁷ ) where: / is thickness [m], static melt layers, g is gravitational acceleration [m.s' ² j, p is density [kg.m ^-3 ] melt, η is the dynamic viscosity [Pa.s] of the melt, and ₀ is the initial radius [m] of the bubble, á is the velocity [ms ^-1 ] of the bubble growth, r is the time [s] required to remove the bubble. enamel.

Thus, if the time τ obtained from equation (17) is lower or comparable to the value obtained for the centrifugation conditions obtained, centrifugation makes no sense. The following table shows the temperatures above which it becomes more advantageous when filling the cylinder V7V ₀ = 0.5 using the rotation before the simple rise of the critical bubble through the layer of glass 7. This is a set of cylinders where 15 h ₀ = 2r ₀ . The values in the table indicate the temperatures at which the critical bubble removal time in the cylinder is the same for r = Os ^-1 and τ-T _opt . The results show, in agreement with the modeling results, that centrifugation is fast and advantageous for model glass above 1400 ° C (for other glasses at bubble growth rates above about 4.10 ' ⁷ m.s ^-1 ).

Table 2

r ₀ [m] 0.10 0.175 0.25 0.375 0.50 0.75 ø [° C] <1350 <1350 1357 1367 1376 1385

2QÍ

Tab. 2 shows the temperatures above which it is preferable to use centrifugation under optimal conditions ω _αρΙ τ _ορι when filling the cylinder I / O _o = 0.5, rather than simply raising the critical bubble through a layer of glass 7 that would _settle in the cylinder without rotation. For cylinder dimensions, h ₀ = 2r ₀ .

Generally speaking, the use of centrifugation for glasses containing fining agents is less advantageous or disadvantageous at low bubble growth rates (at low temperatures) and thick layers of glass 7 in the rotating cylinder.

- t i l ·

Example 2

For centrifugation, a hollow rotating body 7 is provided, in the form of a drum having a radius of 0.25 m and a height of 0.5 m, as shown, for example, in Fig. 4. charge with 35 masses, h% of own shards. The clarified glass is discharged into the processing glass container after cooling to the manual processing temperature. It is necessary to compare two variants of the fining process, differing in the content of the fining agent added to the batch and the temperature. The minimum actual power of the rotating drum was prescribed at 5t / 24h. 1500 s is reserved for filling and emptying the drum. 1U is required. decide whether the case with a lower temperature partially compensated by a higher concentration of finite is applicable. For both strains with prescribed amount of fining agent and at prescribed temperatures, values of average bubble growth rate were obtained:

A. A glass of 0.35 wt% Sb ₂ O ₃ at 1450 ° C has an average bubble growth rate of 15. ά = 5.30x10 ⁷ ms10.

B. A glass with 0.35 wt% Sb ₂ O ₃ at 1420 ° C has an average bubble growth rate ά = 3.20x10 ^-7 ms ^{= 1} .

2ď 1. We find for both variants the respective temperatures corresponding to the same bubble growth rate in the model glass. We use equation (1):

Glass A has a fining temperature of 1408 ° C for the model glass.

Glass B has a temperature of 1383 ° C for the model glass.

2. We select a cylinder filling which should be between 25 and 75% of the cylinder volume.

Select V / V _o = 0.6 and calculate the average thickness of the glass layer in the cylinder according to equation (6). The layer thicknesses are:

Both A and B glass have an average glass layer thickness of 0.092 m.

jtf 3. From equations (7'48) we find the slope or the y-axis segment for the linear dependence of the optimal critical bubble removal time and ₀ = 5x10 ' ⁵ maz and equation (2) the optimal critical bubble removal time, τ,:

For glass A, τ is 731 s.

For glass B the value τ _{ιιρι is} 1326 s.

35 (

4. _{Find the} appropriate values ω _ορι , the optimal angular velocity of the cylinder rotation. First of all, from the equations (13-14) we determine the slopes and the y-axis segments of the linear dependence between ω τ and the cylinder filling V! V ₀ and from equation (3) calculate the product ω _ορ , τ _ορ1 .

For glass A, the value obtained is ω _ηρι τ _ορι 17900 rad. Using the value 5 τ = 731 s we get the value of the cylinder speed ω _ηρι = 24,5 rad.s' ¹ .

For glass B the value obtained is ω _ιιρΙ τ _ορι 40070 rad. Using the value r _{n / 7} , = 1326s we get the value ω _ορι = 30.2 rad.s' ¹ .

5. Find out the real power of the fining drum. One batch of clarified glass at a given cylinder size and filling of 0.6 is 135.4 kg of glass. At the required power of 5t / 24h it is necessary to clear 37 batches of glass in 24 hours.

Result:

For glass A, the optimum fining time is 731 s, the time reserved for filling and emptying the cylinder is 1500 s, the total time for one batch is 2231 s. To make 37 batches, 82 547 s is required. h is slightly exceeded to 5.23 t / 24 h, thus a slight margin is created.

For glass B, the optimum fining time is 1326 s, the total time for a single batch is 2826 s. For 37 batches, 104,562 s is required, the daily drum power would only be 4.13 t / 24 h. therefore, it does not meet the required 5t / 24 hour centrifuge performance condition and therefore it is necessary to use glass A and its associated centrifugation conditions.

Example 3

For example, the glassworks has 2 rotating cylinders (not shown) having an inner radius r ₀ 0.5 m and an internal height h ₀ being 1 m. The glass produced is a container glass and it is assumed that two cylinders will take glass 7 from the continuous furnace and to deliver the clarified melt alternately to a glass container which, after cooling to the processing temperature, will be machine-formed. A total melting capacity of 70 t / 24h is prescribed for the whole furnace, the permissible power tolerances are 3J / 0.

30. The maximum attainable temperature in the cylinder is 145 ° C, which is a temperature sufficient to effectively function the sodium sulfate fining agent. The average bubble growth rate measured at 1450 ° C in the laboratory was ά = 9x10 ⁷ / zz · ¹ . Filling the cylinder from the furnace and emptying it into the working container takes 1800 s. The cylinders are alternately attached to the common outlet opening from the furnace to the inlet and above the working container to which it ripples and discharges. Both cylinders can also be above the hopper at the same time. It is necessary to determine the filling of the cylinder with melt, the optimum speed of rotation of the cylinder and the time of removal of all bubbles, with the assumed radius of the smallest bubble being 5x10 ^-5 m.

2É.

i f

1. Find the appropriate temperature corresponding to the same bubble growth rate in the model and applied glass. We use equation (1). The temperature of the same bubble growth rate in the model glass is 1440 ° C.

2. In the first step, select cylinder filling I / O _o = 0.5 and calculate the average melt thickness in the cylinder according to equation (6). The calculated average thickness of the melt layer in the cylinder is 0.146 m.

3. Find the value of the slope τ versus δσ using equation (7) and calculate τ _πρι ζ of equation ló (2). We get the value of optimal ripple time T / = ^{462 s} ·

4. From the values of the product ύ) _πρί τ _ορί we obtain the value ω _ηρι . First, from equations (13-14) obtain the value of the slope of ω ^ τ and cylinder charging / O and _a section on the vertical axis, then from equation (3) calculate the appropriate value of _ω ηρι τ ορι under the conditions equal to 9575 rad. At the known value τ we calculate the optimum cylinder speed from _n ω _ηρι τ _ορι ® _npl = W radf ·

5. The value of one batch of clarified glass per cylinder is 902.7 kg of glass and this quantity is obtained in 462 s. The total time required for loading, clarifying and pouring one batch of glass is

2.6- then 1800 + 462 = 2262 s. Thus, one cylinder is able to clear 86400/2262 = 38.2, ie 38 batches of glass, ie 34.3 t / 24 h in 24 h ^ 4. the kiln has a finishing power of 68.6 t / 24 h. With a tolerance of 3%, the power is sufficient and the conditions A of the centrifugal fining can be accepted.

Example 4

It is necessary to remove bubbles from 10 kg of special borosilicate glass in the rotating body 7, for example in a small cylinder with a radius of 0.1 m and a height of 0.2 m. The glass in this cylinder is first melted from raw materials and then clarified. The cylinder is also used to remove bubbles from already melted glasses, where it is dosed into it shards of unclear glass. Laboratory measurement at 1480 ° C revealed an average value of bubble growth rate ά = 1,5x10 ' ⁶ m. This is a one-time preparation of quality glass.

The following procedure is chosen:

1. Find the appropriate temperature corresponding to the same bubble growth rate in the model and applied glass. We use equation (1). Temperature of the same rate of bubble growth in the model

M glass is 148 ° C.

2. Calculate the filling of the cylinder for a given mass of glass in the given cylinder: V / V _o = 0,7.

3. According to equation (6), the calculated average melt thickness in the cylinder is 0.045 m.

IQ

4. _{Find the slope} value τ _ηρι versus σ using equation (7) and calculate τ from equation (2). We get the value τ _ορΙ = 90 s.

5. From the values of the product ω _πρΙ τ _ορι we obtain the value ω. First we get '15 from equations (13-, 14). the value of the slope of the dependence between ω τ and the filling of the cylinder V7V ₀ and the section on the y-axis and then from equation (3) we calculate the corresponding value ω _ορι τ _ορι equal to _{5973 rad} under given conditions.

With the known value τ _ορι we calculate from the product ω _{Ι) ρ} , τ _ορ , the value ω _ορ , = 66.4 s' ¹ .

The optimum rotational speed of rotation of the rotating body 1 is thus 66.4 s ^-1 , which corresponds to 19.09 revolutions per second. The molten glass 7 is free of bubbles in less than 2 minutes, specifically in 90 seconds.

Example 5

We have to find the temperature of operating a rotating cylinder with a radius of 0.3 m with a height of 0.4 m, 2Š, which should have a power of 7.5 t / 24h of sodium potassium crystal glass. Filling and emptying time of the cylinder is 1500s.

The following procedure is chosen:

30, 1. Select the filling value of the cylinder I / O _o = 0.6.

2. Determine the mass of one batch of melt, which in this case is 156 kg of glass.

3. Calculate the number of doses required to achieve 7.5 t / 24 h

7500/156 = 48.1 doses.

'*

4. Calculate the value τ _ορ , = (86400 - 48,1x1500) / 48,1 = 296 s.

2. According to equation (6), we calculate an average melt thickness in the cylinder of 0.11 m.

AND

5. From the equation (2) we calculate the slope value k = 296 / 0.11 = 2691 sm ¹ . This slope corresponds to a model glass temperature of 1460 ° C.

6. The model glass temperature of 1460 ° C corresponds to an average bubble growth rate of 1.17x10 ' ⁶ m.s ^-1 . 10

7. The average bubble growth rate should be measured over a narrow temperature range and a temperature corresponding to an average bubble growth rate of 1.17x10 ^-6 m.s ^{-1 is found} .

Table 3 shows the values of the quantities for the exemplary embodiments 2A, 2B, 3, 4 and 5:

Table 3

Example number r ₀ [m] him [m] V / Vo da / dT [m / s] T [° C] W ₀ pt [rad / s] Tcrit [with] It [WITH] P [t / day] 2A 0.25 0.50 0.6 5,30x10 ' ⁷ 1450 24.5 731 1500 5.23 2B 0.25 0.50 0.6 3.20x10 ' ⁷ 1420 30.2 1326 1500 4.13 1 2x0,50 1.00 0.5 9,00x10 ' ⁷ 1450 20.7 462 1800 68.6 3 0.10 0.20 0.7 1.50x10 ' ⁶ 1480 66.4 90 - 0.01 4 0.30 0.40 0.6 1.17x10 ' ⁶ 1460 26.6 296 1500 7.5

t

In Table 3 it represents:

r ₀ [m] radius of the total interior space V _{o of the} cylindrical rotating body 1, ho [m] the height of the total interior space V _{of the} cylindrical rotating body 1,

Vo [m ³ ] total interior space V _{o of} cylindrical rotating body 7 in m ³ , / [m ³ ] space of cylindrical rotating body 1, filled with glass 7 in m ³ ,

V / V _o ratio of cylindrical rotating body 1 filled with glass 7 to total interior space V _{o of} rotating body 1, da / dr derivative of average velocity [m.s ^-1 ] of bubble radius growth in glass 7, at given temperature in rotating body 1

29.

a is the radius [m] of bubbles in meters, r is the time [s] of bubble growth in a given radius in seconds,

a) _opt optimal angular velocity [rad.s ^-1] rotation of the cylinder in radians per second, the temperature T [° C] molten glass 7 in the rotating body 1 in degrees Celsius;

and. TcritQ time [s] of centrifugal bubble of critical radius 5x10 ' ⁵ m at optimal rotation speed, i

τ ₀ total time [s] of filling and emptying of rotating body in seconds, a

P actual fining performance by tumbling the glass in tonnes per day in the non-rotating body 1 with respect to the time it takes to fill and empty it.

1d. These exemplary embodiments are not limiting, and other variants and combinations of design within the scope of the claims are possible.

Industrial applicability

The solution is suitable for the glass industry, for continuous and especially for discontinuous melting with subsequent glass clarification.

2Ó. The reference-marks ^of ·· '' r. '·' · ^Τ - ' ^r ' body cover 2 body bottom 3 body housing 4 body 1

25 ¹ 5 inlet outlet outlet hole glass axis 8 body 1 direction 9 glass inlet 7

10 direction 10 of glass outlet 7 f

direction of rotation of the body 1

Claims (6)

PATENT CLAIMS An apparatus for clarifying glass by centrifuging, in which the bubbles are separated from the glass (7) by centrifugal force, comprising at least one fining centrifuge comprising an X rotating body (1) of cylindrical shape, with a peripheral shell (4), a bottom (3) and optionally a lid (2) having axial symmetry in the axis (8) of rotation of the rotating body (1), the rotating body (1) being equipped with a glass inlet and outlet means (7), characterized in that the cylindrical rotating body (1) has a. the total interior space V _o , partially filled with glass (7) to be clarified o 1Ó. a V-content connected to the ambient atmosphere to escape bubbles from the glass (7); b. the total interior space V _o , which has a mean radius (r ₀ ) of 0.05 ~ -1.0 m; and the total height h ₀ is 0.1 ^ -1.5 m; wherein the proportion of the content V of the rotating body (1) filled with the glass (7) to be clarified is relative to the total internal content V _{o of the} rotating body (1) in proportion 15 I / O _o = 0.20 U0.80; and c. The rotational speed of the rotating body (1) of the fining centrifuge during centrifugation is in the range of 10-200 rad.s- ¹ . Device according to claim 1, characterized in that the rotating body (1) with the cylindrical casing (4) 20 and the bottom (3) is completely open at the upper edge of the mouth, and this open mouth serves as the inlet opening (5) of the glass (7). ). Device according to claim 1, characterized in that the rotating body (1) with the cylindrical shell (4) and the bottom (3) has at its upper end a solid wall adjoining the housing (4) or is an upper end 25 'provided with a removable lid (2), the upper wall or lid (2) being equipped with at least one reversible closable inlet opening (5) of the glass (7), namely an inlet opening (5) in the axis (8) of the rotating body (1) and / or through inlets (5) outside the axis (8) of the rotating body (1). Device according to claim 1, characterized in that the rotating body (1) with the cylindrical shell (4) 30 has a bottom (3) without an outlet opening (6) of the glass (7). Apparatus according to claim 2, characterized in that the rotating body (1) with the cylindrical housing (4) has a bottom (3) provided with at least one returnable outlet opening (6), namely an outlet opening (6) in the axis ( 8) a rotating body (1) and / or outflow openings (6) off the axis 35 '(8) of the rotating body (1). r * < Device according to one of Claims 1 to 7, characterized in that the rotating centrifuge body (1), its bottom (3), the casing (4) and possibly the lid (2) are made of a refractory ceramic or a refractory metal or alloy , with an outer steel jacket, and optionally, a thermal insulation is located below the steel jacket. Apparatus according to claim 1, characterized in that the clarifying centrifuge is heated by gas or electrically and optionally by microwave heating.