JP5157385B2 - Method for producing glass particulate deposit - Google Patents

Description

  The present invention relates to a method for producing a glass fine particle deposit capable of reducing the frequency of cleaning processing performed between batches and improving the productivity by improving the apparatus operating rate.

As one of the processes for producing a quartz glass preform for optical fiber, there is a process for producing a glass fine particle deposit in which glass fine particles are deposited on a starting substrate.
As a method for producing the glass fine particle deposit, for example, there is a method using an MMD method (multi-burner multilayer attachment method) or an OVD method (external attachment method).
In these manufacturing methods, the glass raw material sprayed from the burner is flame-hydrolyzed in an oxyhydrogen flame, and the produced glass fine particles (soot) are deposited in multiple layers on the starting substrate (target bar).

However, not all of the glass particles generated from the burner are deposited on the starting substrate, and the glass particles that have not adhered to the starting substrate (non-attached soot) float in the chamber of the apparatus. Become.
Normally, since the chamber of the apparatus is equipped with an exhaust mechanism that sucks and collects the atmosphere inside, usually most of the unattached soot floating in the chamber is discharged out of the chamber by the exhaust mechanism. Unattached soot that has not been discharged adheres to the chamber wall surface while floating in the chamber, and accumulates as soot.

  If the soot deposited on the wall surface of the chamber peels off from the wall surface and adheres to the glass fine particle deposit during the deposition process, it may cause bubbles to be generated as a quality defect in the process of converting the deposit into a transparent glass. Therefore, it is necessary to perform timely cleaning processing for removing the soot adhering to the chamber wall surface using a vacuum cleaner or the like before starting the deposition of the next batch.

FIG. 6 shows a work process between batches that incorporates a cleaning process for removing the soot adhering to the chamber wall surface.
That is, in the work between batches, when the deposition of the specified amount of glass particles on the starting substrate is completed (step S101), the deposited glass particle deposit is taken out from the apparatus (step S102), and then the vacuum cleaner. Is used to remove the soot adhering to the chamber wall surface (step S103), and then the starting material of the next batch is set in the chamber (step S104). This is a procedure for starting batch processing (steps S105 and S106).

However, the cleaning process for removing the soot adhering to the chamber wall surface is time-consuming. For example, if the cleaning process is performed for each batch, the productivity is lowered due to a reduction in the operating rate of the apparatus.
Therefore, by adjusting the minimum internal pressure Pmin in the chamber of the apparatus for depositing the glass fine particles on the starting substrate so as to be in the range of −80 Pa ≦ Pmin ≦ −40 Pa with the differential pressure inside and outside the apparatus, the exhaust of unattached soot A manufacturing method has been proposed in which the exhaust efficiency of the mechanism is increased and the frequency of the cleaning process is reduced by suppressing the accumulation of soot on the chamber wall surface (see, for example, Patent Document 1).

JP 2005-247624 A

If the minimum internal pressure Pmin in the chamber is managed as in the manufacturing method of the above-mentioned Patent Document 1, the exhaust efficiency by the exhaust mechanism is increased as compared with the conventional one, so that soot deposition itself on the chamber wall surface can be reduced accordingly. By reducing the frequency of the cleaning process, the operating rate of the apparatus can be improved and the productivity of the glass particulate deposit can be improved.
However, it is impossible to eliminate the soot accumulation on the chamber wall surface. Even if the minimum internal pressure Pmin in the chamber is controlled, the exhaust efficiency of unattached soot will fluctuate if the supply amount of raw material gas to the burner fluctuates. It cannot be accurately grasped.
For this reason, the proper implementation timing of the cleaning process is missed, so that the soot separated from the chamber wall surface adheres to the glass particulate deposit during the deposition process. There has been a problem that it is difficult to stably produce a high-quality glass fine particle deposit that does not cause the occurrence of water vapor.

  An object of the present invention is to solve the above-mentioned problems, and reduce the frequency of cleaning processing for removing soot adhering to the chamber wall surface, and improve the productivity of the glass particulate deposit by improving the operating rate of the apparatus. In addition, there is no inconvenience that the soot peeled off from the chamber wall due to missing the proper timing of the cleaning process does not adhere to the glass particle deposits that are being deposited, and stable production of high-quality glass particle deposits is possible. An object of the present invention is to provide a method for producing a glass particulate deposit that can be produced.

(1) In order to solve the above-described problems, a method for producing a glass fine particle deposit according to the present invention includes a burner that generates glass fine particles toward a starting substrate, and glass fine particles that are not deposited on the starting substrate. In the method for producing a glass fine particle deposit by an MMD method or an OVD method in an apparatus having an exhaust mechanism for exhausting, the exhaust mechanism out of the glass fine particles not deposited on the starting substrate is exhausted. When 50% or more and less than 70% of the fine glass particles not deposited on the starting substrate are exhausted by the exhaust mechanism, the frequency of the cleaning process is reduced to 1/2 to 1/4 (times / When 70% or more of the glass fine particles not deposited on the starting substrate are exhausted by the exhaust mechanism, the frequency of the cleaning process is set to 1/5 to 1/10 (times / bar It is to set the switch).

(2) Further, in the method for producing a glass particulate deposit described in (1) above , clean air is introduced into the apparatus from an air introduction section provided around the burner, and the flow rate of clean air in the apparatus May be maintained at 1 m / s or more.

(3) Moreover, the manufacturing method of the glass fine particle deposit as described in said (1) or (2) is good also considering the said air introduction part as a positive pressure.

(4) Moreover, in the method for producing a glass fine particle deposit according to any one of (1) to (3), the surface roughness (Ra) of the muffle inner surface is set to a rough surface of 0.4 μm or more. Also good.

According to the method for producing a glass particulate deposit according to the present invention, the calculation of the exhaust rate means that the amount of unattached soot actually deposited on the chamber wall of the apparatus is calculated, and the actual soot on the chamber wall surface is calculated. Accumulate the amount of deposition.
And when the exhaust rate is high, conversely, the amount of unattached soot actually deposited on the chamber wall of the apparatus is reduced, so it is extremely reasonable to set the cleaning processing frequency low at that time. .
That is, the method for producing a glass particulate deposit according to the present invention reduces the frequency of the cleaning process because the cleaning process is thinned out when the amount of unattached soot deposited on the chamber wall of the apparatus is actually reduced. Thus, the productivity of the glass particulate deposit can be improved by improving the operating rate of the apparatus.
In addition, since it is estimated that the amount of unattached soot accumulated on the chamber wall reaches a predetermined amount and the cleaning process is performed, the proper execution timing of the cleaning process is not missed without missing the proper execution timing of the cleaning process. It is possible to avoid the inconvenience that the soot peeled off from the wall surface of the chamber adheres to the glass fine particle deposit during deposition, and a high-quality glass fine particle deposit can be stably produced.

Hereinafter, a preferred embodiment of a method for producing a glass fine particle deposit according to the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic configuration diagram of an embodiment of an apparatus for carrying out a method for producing a glass particulate deposit according to the present invention, and FIG. 2 sets the frequency of cleaning processing in the method for producing a glass particulate deposit according to the present invention. It is a flowchart of a process.

A manufacturing apparatus 1 shown in FIG. 1 is an apparatus that manufactures a glass fine particle deposit by the MMD method. A starting substrate (start glass rod) 5 disposed in a chamber 3 that is a reaction vessel includes traverse apparatuses 7a and 7b. It is connected to support rods 8a and 8b extending from the shaft so as to be able to reciprocate in the axial direction.
A large number of burners 11 a to 11 h are arranged on one side of the chamber 3 so as to be arranged at regular intervals in the longitudinal direction of the starting substrate 5, and the atmosphere in the chamber 3 is negatively pressurized on the other side of the chamber 3. Exhaust mechanisms (exhaust pipes) 13a, 13b, and 13c for exhausting unattached soot floating in the chamber 3 to the outside by suction and exhaust are disposed.

Moreover, the gas supply path 15 which supplies raw material gas required for the production | generation of glass particulates, general gas, etc. is connected to each burner 11a-11h.
Each burner 11a-11h injects the glass raw material gas supplied from the gas supply path 15 toward the starting base material 5, and makes the glass fine particle (soot) produced | generated by the flame hydrolysis of the glass raw material gas a starting base material 5 is deposited in multiple layers on the outer periphery of 5 to form a glass particulate deposit 17.

An air introduction part 19 communicating with the inside of the chamber 3 is defined around each of the burners 11a to 11h on one side of the chamber 3, and an air supply pipe 21 for introducing clean air is provided in the air introduction part 19. Is connected.
The air introduction unit 19 introduces clean air at a predetermined flow rate into the chamber 3 from the peripheral portions of the burners 11a to 11h, thereby generating an air flow that makes unattached soot easily reach the exhaust mechanisms 13a, 13b, and 13c. Form in.

  A plurality of pressure detection positions 23 for measuring the pressure in the introduction section are set in the air introduction section 19, and the pressure is measured by a pressure detection means (not shown) at these pressure detection positions 23.

A plurality of pressure detection positions 25 are set so as to be located between the exhaust mechanisms 13 a, 13 b, 13 c disposed on the other side of the chamber 3 and the glass particulate deposits 17.
At these pressure detection positions 25, the pressure is measured by pressure detection means (not shown), and the suction pressures by the exhaust mechanisms 13a, 13b, 13c are set so that each pressure detection position 25 is maintained at a predetermined negative pressure. Adjusted.

  The manufacturing method of the glass particulate deposit carried out by the production apparatus 1 forms glass particulate deposits 17 by spraying glass particulates from a large number of burners 11a to 11h toward the starting substrate 5 as shown in the figure. Although the manufacturing method is based on the so-called MMD method, as shown in FIG. 2, the ratio (A%) exhausted by the exhaust mechanisms 13a, 13b, 13c out of the total amount of glass fine particles not deposited on the starting substrate 5 is calculated. However (step S201 in FIG. 2), the higher the exhaust ratio A, the lower the frequency of the cleaning process for removing the soot adhering to the inside of the apparatus.

The exhaust ratio A is calculated at the end of the batch as follows.
First, the total weight w1 of the amount of glass fine particles produced is calculated from the raw material input from each of the burners 11a to 11h. Further, the weight w4 of only the glass fine particle portion of the glass fine particle deposit 17 is measured by subtracting the weight w3 of the starting substrate 5 from the total weight w2 of the glass fine particle deposit 17. Next, the soot adhesion weight w5 of the chamber 3 is measured.
From the above measured weights,
Exhaust ratio A = (w1- (w4 + w5)) / (w1-w4) (1)

And the frequency of the cleaning process in subsequent batches is adjusted according to the exhaust ratio A calculated by the above equation (1).
When the exhaust ratio A is 50% or more and less than 70%, the frequency of the cleaning process is set to 1/2 to 1/4 (times / batch) as shown in Step S210 and Step S211 of FIG. This means that the cleaning process is thinned out once every 2 to 4 batches.
When the exhaust ratio A is 70% or more, as shown in steps S220 and S221 of FIG. 2, the frequency of the cleaning process is set to 1/5 to 1/10 (times / batch).
Further, when none of the above steps S210 and S220 correspond (that is, the exhaust ratio A is less than 50%), as shown in step S231 in FIG. Set to times.

  The frequency of the cleaning process as described above is based on the results of examining the correlation between the number of bubbles generated when a glass particle deposit produced in the past is converted to transparent glass, the exhaust ratio, the frequency of the cleaning process, and the like. .

  Since the non-operating time of the apparatus is inversely proportional to the frequency of the cleaning process as shown in FIG. 4 described later, the non-operating time of the apparatus is reduced by reducing the frequency of the cleaning process. The productivity of the glass particulate deposit can be improved by improving the operating rate.

  In addition, clean air is introduced into the apparatus from the air introduction unit 19, but the flow rate of clean air in the apparatus is maintained at 1 m / s or more. As a result, unattached soot floating in the chamber 3 efficiently flows to the exhaust mechanisms 13a, 13b, and 13c.

  Furthermore, desirably, the amount of clean air introduced from the air supply pipe 21 is set so that each pressure detection position 23 is maintained at a predetermined positive pressure (that is, the air introduction portion 19 is maintained at a positive pressure). adjust. Thus, by setting the air introduction part 19 to a positive pressure, it is possible to prevent external foreign matters from being mixed into the chamber 3 or the air introduction part 19.

More preferably, the surface roughness (Ra) of the inner surface of the muffle, which is the inner surface of the chamber 3, may be 0.4 μm or more.
The inventors of the present application are able to easily peel off the glass particles adhering to the inner surface of the muffle, change the surface roughness Ra of the inner surface of the muffle, the number of bubbles generated in the glass base material made into transparent glass, and the surface roughness of the inner surface of the muffle. The relationship with Ra was investigated. Then, when the surface roughness Ra of the muffle inner surface was 0.4 μm or more, the generation of bubbles in the base material was not detected. When the surface roughness Ra of the muffle inner surface is less than 0.4 μm, the generation of bubbles in the base material rapidly increases as the surface becomes smooth. The smoother the inner surface of the muffle, the lower the adhesion strength of the soot to the inner surface of the muffle, and it is thought that the soot peeled off from the inner surface of the muffle during the base material deposition process adhered to the base material. By using a device whose inner surface roughness Ra is roughened to 0.4 μm or more, it is possible to prevent generation of bubbles due to adhesion of peeled soot to the glass base material, and high reliability. Deposition processing can be performed.

  According to the method for producing a glass particulate deposit in the above embodiment, the calculation of the exhaust ratio is the calculation of the amount of unattached soot w5 that is actually deposited on the chamber wall of the apparatus. It is possible to accurately grasp the actual soot accumulation amount w5.

And when the exhaust rate is high, the amount of unattached soot w5 actually deposited on the chamber wall of the apparatus is reduced, so it is extremely reasonable to set the cleaning processing frequency low at that time. is there.
That is, the manufacturing method of the glass particulate deposit according to the present embodiment thins out the cleaning process when the amount of unattached soot w5 actually deposited on the chamber wall of the apparatus is reduced. The productivity of the glass fine particle deposit can be improved by improving the operating rate of the apparatus.

  In addition, since it is estimated that the amount of unattached soot w5 deposited on the chamber wall reaches a predetermined amount and the cleaning process is performed, the proper execution timing of the cleaning process is not missed without missing the proper execution timing of the cleaning process. Therefore, it is possible to avoid the inconvenience that the soot peeled from the wall surface of the chamber adheres to the glass fine particle deposit during deposition, so that a high-quality glass fine particle deposit can be stably produced.

Next, the glass fine particle deposits produced in the specific examples and comparative examples of the above embodiment will be described.
In each of the following Examples 1 to 18 and Comparative Examples 1 to 8, the manufacturing apparatus 1 uses a core glass for an optical fiber having a diameter of 40 mm and a length of 2100 mm as the starting substrate 5. The average amount of glass raw material (SiCl ₄ ) charged into the eight burners 11a to 11h is 5 SLM / piece. Then, when the amount of glass fine particles deposited on the starting substrate 5 reaches 100 kg, the glass fine particle deposit 17 is taken out as the manufacture of the glass fine particle deposit 17 is completed.

In each of Examples 1 to 18, the exhaust mechanism 13a, 13b, or 13c was adjusted so that the measured value at each pressure detection position 25 shown in FIG. Adjust the exhaust.
In each of Examples 1 to 18 and Comparative Examples 1 to 8, 60 glass fine particle deposits 17 are manufactured, and the 60th glass fine particle deposit 17 is made into a transparent glass and the number of bubbles is measured.

Further, (total amount of glass particulates exhausted by exhaust mechanisms 13a, 13b, 13c attached to the apparatus) / (total amount of glass particulates not deposited on starting substrate 5) × 100 = exhaust ratio A (%), The number of batches until the cleaning process is performed is B, the flow rate of clean air in the chamber 3 is C (m / s), and the pressure of the air introduction unit 19 at each pressure detection position 23 shown in FIG. 1 is D (Pa). The value of the surface roughness Ra of the inner surface of the muffle is E (μm), and the number of bubbles F of 1 mm or more in the glass base material obtained by transparent vitrification of the 60th glass fine particle deposit produced in each example and comparative example ( Piece / base material) and average equipment stop time G (min) between batches.
Tables 1 to 3 below show the measurement results of the examples and comparative examples.
Hereinafter, the measurement results of each table will be described.

In Table 1, Examples 1 to 3 are cases where the exhaust ratio A is 50%, and the frequency of the cleaning process is set according to the flow shown in FIG. Moreover, in each of these Examples 1 to 3, the flow rate C of clean air is 1 m / s, the pressure D of the air introduction part 19 is 0.60 Pa, and the surface roughness E of the inner surface of the muffle is 0.56 μm.
In each of Examples 1 to 3, the number of bubbles F in the measured base material was 0 or 1 (pieces / base material), and a good base material with few generated bubbles was obtained.

On the other hand, in Comparative Example 1, the frequency of the cleaning process was performed once per batch, and other manufacturing conditions were the same as those in Examples 1 to 3. Compared with Examples 1 to 3, good results were confirmed for the number of bubbles F, but the average equipment stoppage time G is longer than that of Examples 1 to 3 because the cleaning frequency is high.
That is, in Examples 1 to 3, the average equipment stop time G (minutes) is shortened by reducing the cleaning frequency as compared with Comparative Example 1 in which the cleaning process is performed for each batch, and the operating rate of the apparatus Was confirmed to be improved.

  In Comparative Example 2, the frequency of the cleaning process is set to once in 5 batches without following the flow shown in FIG. 2, and other manufacturing conditions are the same as in Examples 1 to 3. Yes. In this case, the frequency of the cleaning process is excessively lowered, and the number of bubbles F in the measured base material becomes 10 (pieces / base material), and the effectiveness of the flow shown in FIG. 2 can be confirmed. It was.

In Table 1, Examples 4 to 6 are cases where the exhaust ratio A is 69%, and the frequency of the cleaning process is set according to the flow shown in FIG. Further, in each of these Examples 4 to 6, the flow rate C of clean air is 1.38 m / s, the pressure D of the air introduction part 19 is 0.83 Pa, and the surface roughness E of the muffle inner surface is 0.56 μm.
In each of these Examples 4 to 6, the number of bubbles F in the measured base material was 0 or 1 (pieces / base material), and a good base material with a small number of generated bubbles was obtained. .

In Comparative Example 3, for comparison with Examples 4 to 6 above, the frequency of the cleaning process was set to once in 5 batches without following the flow shown in FIG. The conditions are the same as in Examples 4-6. In this case, the number of bubbles F in the measured base material is 9 (pieces / base material) due to an excessive reduction in the frequency of the cleaning process.
As in the case of Comparative Example 2, the effectiveness of the flow shown in FIG. 2 was confirmed.

In Table 1, Examples 7 to 12 are cases where the exhaust rate A is 90%, and the frequency of the cleaning process is set according to the flow shown in FIG. In each of these Examples 7 to 12, the flow rate C of clean air is 1.8 m / s, the pressure D of the air introduction part 19 is 1.08 Pa, and the surface roughness E of the muffle inner surface is 0.56 μm.
In each of Examples 7 to 12, the number of bubbles F in the measured base material was 0 or 1 (pieces / base material), and a good base material with a small number of generated bubbles was obtained.
Moreover, since the frequency of the cleaning process is significantly reduced as compared with Comparative Example 1, the average facility stop time G is significantly shortened.
Comparative Example 4 is the case where the exhaust ratio A is 90%, which is the same as in Examples 7 to 12, but the frequency of the cleaning process is reduced to once in 11 batches without following the flow shown in FIG. The other manufacturing conditions are the same as in Examples 7-12. In this case, the frequency of the cleaning process is excessively lowered, and the number of bubbles F in the measured base material becomes 10 (pieces / base material), and the effectiveness of the flow shown in FIG. 2 can be confirmed. It was.

FIG. 3 summarizes the correlation between the number of batches B to be cleaned and the number of bubbles F generated in the base material for each exhaust efficiency A in each example and comparative example shown in Table 1. When the exhaust efficiency is 50% and the exhaust efficiency is 69%, the same tendency is shown. When the number of batches B exceeds 4, the number of bubbles F tends to increase rapidly, as shown in FIG. When 50% or more and less than 70% of the glass fine particles not deposited on the starting substrate are exhausted by the exhaust mechanism, the frequency of the cleaning process is set to 1/2 to 1/4 (times / batch). This was confirmed to be effective in reducing the number of bubbles F.
Further, when the exhaust efficiency is 90%, when the number of batches B exceeds 10, the number of bubbles F tends to increase rapidly. As shown in FIG. When 70 to 90% of the air is exhausted by the exhaust mechanism, it is effective to reduce the number of bubbles F by setting the frequency of the cleaning process to 1/5 to 1/10 (times / batch). Was confirmed.

  FIG. 4 summarizes the correlation between the number of batches B to be cleaned and the non-operating time of the apparatus in each example and comparative example shown in Table 1, and the non-operating time of the apparatus is the frequency of the cleaning process. Since the relationship is inversely proportional, it was confirmed that reducing the frequency of the cleaning process is effective in improving the productivity of the glass particulate deposit by improving the operating rate of the apparatus.

Examples 13 to 16 and Comparative Examples 5 and 6 shown in Table 2 demonstrate the usefulness of regulating the surface roughness of the muffle inner surface in the above embodiment.
In Examples 13 to 16 and Comparative Examples 5 and 6, the exhaust rate A is 50%, and the frequency of the cleaning process is set to once every two batches according to the flow shown in FIG.
Further, the flow rate C of clean air is 1 m / s, and the pressure D of the air introduction part 19 is 0.60 Pa.

  However, as for the surface roughness E of the inner surface of the muffle, Examples 13 to 16 are set to a value of 0.4 μm or more according to the condition setting of the embodiment, but Comparative Examples 5 and 6 are those of the embodiment. It is set to 0.4 μm or less that deviates from the condition setting. As a result, in Examples 13 to 16, the measured number of bubbles F in the base material is 0 (pieces / base material) as shown in FIG. 5, and the number of generated bubbles is small and good. In the case of Comparative Examples 5 and 6, the number of bubbles F in the measured base material was 10 or 20, and the surface roughness E of the inner surface of the muffle was 0.4 μm or less. As a result, it became clear that the quality deteriorated, and the usefulness of setting the surface roughness of the inner surface of the muffle shown in the above embodiment to 0.4 μm or more could be confirmed.

Examples 17 and 18 and Comparative Examples 7 and 8 shown in Table 3 demonstrate the usefulness of regulating the flow rate C of clean air and the pressure D of the air introduction part 19 in the above embodiment. .
In these Examples 17 and 18, when the exhaust ratio A is 50%, the frequency of the cleaning process is set to once every two batches according to the flow shown in FIG.
The flow rate C of clean air is 1 m / s, and the pressure D of the air introduction part 19 is a positive pressure of 0.60 Pa.
On the other hand, in Comparative Example 7, when the exhaust rate A is reduced to 48% by reducing the flow velocity C to 0.96 m / s, the frequency of the cleaning process follows the flow shown in FIG. The place that must be once in one batch is set to once in two batches.

As a result, it was confirmed that the number of bubbles F in the measured base material was 5 and the quality was deteriorated by making the flow rate C of clean air and the cleaning frequency out of regulation.
Comparative Example 8 is Example 17 except that the exhaust ratio A is 50%, and the pressure D of the air introduction portion 19 is set to a negative pressure (−0.50 Pa) that is outside the regulation of the embodiment. 18 under the same manufacturing conditions, but due to the influence of the negative pressure D of the air introduction part 19, the measured number of bubbles F in the base material becomes 10 (pieces / base material), and air is introduced. The usefulness of setting the pressure D of the part 19 to a positive pressure was confirmed.

  In the above embodiments and examples, the case where the glass fine particle deposit is manufactured by the MMD method has been described. However, the manufacturing method of the present invention is also applicable when the glass fine particle deposit is manufactured by the OVD method. It is.

It is a schematic block diagram of one Embodiment of the apparatus which enforces the manufacturing method of the glass fine particle deposit body concerning this invention. It is a flowchart of the process which sets the frequency of the cleaning process in the manufacturing method of the glass particulate deposits concerning the present invention. It is a graph which shows the correlation with the cleaning frequency in the Example of this invention, and the bubble number in a base material. It is a graph which shows the correlation with the cleaning frequency in the Example of this invention, and the non-operation time of an apparatus. It is a graph which shows the correlation with the muffle inner surface precision in the Example of this invention, and the number of bubbles in a base material. It is explanatory drawing of the operation | work between batches at the time of manufacturing a glass particulate deposit.

Explanation of symbols

DESCRIPTION OF SYMBOLS 3 Chamber 5 Starting base material 11a-11h Burner 13a, 13b, 13c Exhaust mechanism 15 Gas supply path 17 Glass fine particle deposit body 19 Air introduction part 21 Air supply pipe 23 Pressure detection position 25 Pressure detection position

Claims (4)

A glass particulate deposit is produced by the MMD method or the OVD method in an apparatus having a burner for generating glass particulates toward the starting substrate and an exhaust mechanism for exhausting the glass particulates not deposited on the starting substrate. In the method for producing a glass particulate deposit,
Calculate the proportion of glass particles not deposited on the starting substrate that are exhausted by the exhaust mechanism,
When 50% or more and less than 70% of the glass fine particles not deposited on the starting substrate are exhausted by the exhaust mechanism, the frequency of the cleaning process is set to 1/2 to 1/4 (times / batch), When 70% or more of the fine glass particles not deposited on the starting substrate are exhausted by the exhaust mechanism, the frequency of the cleaning process is set to 1/5 to 1/10 (times / batch). A method for producing a glass particulate deposit.   2. The glass fine particles according to claim 1, wherein clean air is introduced into the apparatus from an air introduction portion provided around the burner, and the flow rate of the clean air in the apparatus is maintained at 1 m / s or more. A method for producing a deposit.   The method for producing a glass particulate deposit according to claim 1 or 2, wherein the air introduction part is set to a positive pressure. The method for producing a glass particulate deposit according to any one of claims 1 to 3, wherein the surface roughness (Ra) of the inner surface of the muffle is a rough surface of 0.4 µm or more .

Images