dark glasses which absorb the heat radiation,the side walls and the bottom of the melt-zone may be heat Insulated. The superstructure, apart from the crown, is usually made from stlllmanite refractories, but for melting glasses with a hlgh boron-oxide content, it is recommended to use silica. The crown is made from silica and ts Insulated by a layer of quartz sand,over which is laid a layer of refractory high-silica heat Insulation, heat Insulating slabs from tripoHte, and finally a layer of insulating cement. The temperature differential In the furnace crown Is usually over 1400~ Small furnaces of this type, having no heat'-reeuperator or regenerator arrangements, are economical and produce good quality glass. Certain difficulties were encountered In operating the earlier furnaces, but these were overcome b y deepening the bath, using heat insulation to prevent the glass from cooling in certain parts of the furnace and finally in melting dark glasses containing iron by using auxiliary electrfcal heating (it is recommended to Insert the electrodes through the walls at some distance from the tank bottom). The most intensive corrosion of the refractories is observed in the refining zone, especlally near the glass surface, but blocks which have disintegrated tn these places can be replaced without stopping the furnace. The bottom and superstructure disintegrate comparatively little and last for the furnace campaign. Small furnaces of this type are successfully used for melting different glasses, including colorless, light green, dark green and amber bottle glass, technical glasses with a high lead and barium content, heat-resisting borosilicate, and alkali-free borosilicate glasses for glass-fiber making. The building of these furnaces continues. The cost of furnaces without reeuperators and regenerators is estimated as about half of the ordinary regenerative or recuperative tank furnaces of the same output. Repair costs are also somewhat lower.

SOME G L A S S - M E L T I N G J. T o o l e y , 71 N o .

PROBLEMS

C. Bradley,

2, 1 9 5 8 ,

pp.

A. L y l e

e t

al,,

"Ceramic

Age"

(USA)

34-38

A conference was held in the USA on glass-melting problems, at which the possibilities of employing new types of glass-melting furnaces, as well as certain essential practical problems connected with glass-a--nelting techniques in furnaces of conventional design, were discussed. The main reaslon for the low efficiency of conventional tank furnaces is the combination in them of four distinct processes (melting,homogenization, refinement and cooling). It would be possible, in separate batchmelting furnaces, to melt the batch at a lower temperature, with substantially lower heat losses, and also with a lower pulverization and evaporation of the batch constituents. It is conceivable to split the glass batch into two or three relatively low-melting mixtures chosen from the composition, after studying the structural diagrams of the requisite systems. After the separate melting of each of these mixtures at 800-900~ the entire batch as a whole could then be melted at a temperature of not more than 1200"C, i.e., at temperatures 300-400"C lower than those at which glass is being currently melted. At the same time it would be possible to substantially accelerate the glass melt. The glass should be homogenized and refined in a separate chamber in vacuum in order to quickly remove gas bubbles from the molten glass. It may finally be feasible to refine the glass centrifugally and, after rapid cooling,to pass it on to the forehearth. Completing the different glass melting stages in individual plant units would permit a reduction in heat Input for the melt, re&ice refractory costs, and produce glass of improved quality. Cheap refractory materials, such as fused-quartz block, could be used for glass melting, compensating for their dissolution in the glass by proportioning the batch composition.

676

It would probably be possible to melt the glass In the batch-melting furnace without using refractory materials at the glass-surface level, by laying the top row o f the tank with briquettes made from the batch. It is also feasible to arrange the melting of glass in a shaft furnace using granulated batch from very finely crushed raw materials. This ground batch could be heated up to 700-800"C by waste gases in the top part of the furnace without risk of its stratification. The shaft furnace should be divided Into several zones, temperature differing with height, and to tap the completely melted glass from the bottom. Batch briquetting or granulation, though not economical In conventional tank furnaces, would be fully justified in a shaft furnace. Centraltzedbatch preparation in special factories could be justified for glass melting in small local factories by the conventional methods. The counter flow tank furnaces without regenerators or recuperators, wblch are already In operation In a number of enterprises, show great promise for this kind of factory. The granulated batch should be charged Into these furnaces already In a partially melted condition, so that it should only remain for them to refIne and homogenize the glass and to produce the glass ware. In their discussions on individual essential glass melting problems, the members of the ,conference failed to reach agreement. Thus, certain specialists considered that further increases in the glass-resistance of refractory materials was unrealistic, because their disintegration product~ only dissolve with great difficulty in the glass. Other glass engineers,on the other hand, demanded further increases in the endurance of refractory m a terials so as to be able to provide temperatures of up to 200"C for the glass melt, and to raise glass yields still further. In the discussion on the problem of convection currents, there were also some differences of opinion. Some regarded that convectional currents do more harm than good, while others postulated that currents in general play no very important part in the tank furnace because in the modern furnace, owing to high glass yields, Its currents are governed by the heavy pull, which in principle is different from the convection currents. Glass melting has recently been substantially accelerated through the introduction of fluoride Into the glass composition, which is added into the batch as calcium fluoride in place of part of the lime. By introducing 0.3% fluoride into the glass, the homogenization and refinement are substantially accelerated. Larger additions of fluoride, however, somewhat impair homogenization. Higher melting temperatures in conventional types of tank furnaces ensure a more rapid melt, homogenizatlon and refinement, but raise heat losses, intensify refractory corrosion and the vaporization of the glass c o m ponents. The latter can be prevented tO a considerable degree by heating the glass by electricity. Electric glass melting and auxiliary electric heating in flame furnaces facilitate the raising of melting temperature in excess of those which are currently attained in flame furnaces.

ELECTRIC

TANK

FURNACES

A. E r l k s o n , A m e r i c a n 1958,

pp.

WITH

Ceram.

AUXILIARY

Soc.

GAS

FIRING

Bull (USA),No.

4,

177-1'/8

Three electric-bridge tank furnaces are in operation at one of the American glass-melting plants. When it became necessary to Increase the total yield from these furnaces from 32 to 45 tons per diem, it was decided to employ auxiliary gas heating in the melt-end. Heating by electricity alone, the tank furnaces had no crown,but the glass in the melt-end was completely covered by a thick layer (0.35-0.45 m) of melting batch. The reconstruction of the furnaces involved the suspension of a roof over them, while aworm-batch feed and two burners capable of operating on natural gas or liquid fuel were installed in the rear wall of the fired bath.

677