Advances in Industrial Physics in the Glass Industry

E. K. Sullivan¹), New York

For the research chemist working in the glass industry, it soon becomes obvious that, in order to solve a large part of the problems that arise, training in physics is required. The useful properties of glass, except for resistance to weathering and to other chemical destruction, are physical properties, and many problems of glass manufacture are physical problems. Accordingly, although glass manufacture is usually classed as a chemical industry, in our²) admittedly small staff there are twice as many physicists as chemists.

We believe that for a physicist in the glass industry a good knowledge of the fundamentals of physics and research experience are required, while specialization in molecular physics, or still more in glass, is by no means necessary.

The industrial physicist at a glass plant, as at any other plant, has the following functions: supervision of product quality, improvement of methods of production, and development of the production of new types of products. In addition, he provides technical assistance in patent and legal matters, and also in an innumerable multitude of production problems. If, for example, glass plates are being produced that serve as a building material, physicists are asked about the transparency, thermal conductivity, and sound conductivity of the glass, and its resistance to load, heat, and weathering. The physicist is required to indicate what the surface of the glass should be like so that it will reflect to the required extent the heat and light radiation falling on it from outside, so that visibility through such glass will be sufficiently poor or good, etc.

If window glass is being produced that absorbs heat, the physicist must calculate the amount of energy saved in air-conditioning when such glass is used. The physicist may also be required to make calculations for all sorts of objects, from balloons for electromagnets to lenses for signal beacons, from bottles to glass-melting furnaces.

A furnace for continuous glass melting is a bath into one end of which the batch is charged, while from the other end there is continuous delivery of molten glass. The physicist studies in the glass melt the direction of convection currents, which seek to mix partially molten material with fully molten material and thus slow down the melting process, and shows how the proper distribution of temperature on the surface of the glass eliminates such mixing. If the melted glass contains bubbles of gas, particles of undissolved sand, or crystallization material, the physicist provides

¹) Journal of Applied Physics, 1937. Translated by A. L. Komarova.
²) The author works at Corning Glass Works, Corning, New York.

help in correcting these defects. Such properties of vital importance for glassmaking as, for example, the transparency of the glass produced, the dependence of the release of gas bubbles on the size of these bubbles and on the viscosity of the glass, and consequently also on temperature, also fall within the activity of the physicist.

Perhaps the most valuable contribution to the problem of creating a refractory material for glass-melting furnaces—the most difficult problem in the glass industry—was made by a physicist.

A glass-melting furnace is built of bricks 12 dm thick, with a base area of \(2 \times 3\) feet. These rather porous clay bricks dissolve in the molten glass, especially in its upper layers, and are destroyed within a few months. Every year production is usually stopped for several weeks in order to cool the furnaces and replace the worn bricks.

Corundum—the crystalline form of alumina—and mullite—a crystalline compound of alumina with silica—are known for their resistance to the destructive action of glass. In the search for a material for furnace bricks, the physicist arrived at the idea of smelting aluminum silicate in an electric furnace and casting it in molds, just as is done in casting iron. The bricks obtained are 2–3 times more resistant to the destructive action of glass than clay bricks, and can be heated to a higher temperature without softening. The use of such bricks lengthened the life of furnaces, noticeably improved the average quality of glass owing to a reduction in the number of inclusions obtained in large quantity when clay bricks are used, and increased productive capacity by making it possible to raise the melting temperature. Accordingly, the cost of melting and the depreciation charges per ton of glass decreased, as did the required capital investments. The high appraisal of the new type of brick by glass builders found expression in the continuous growth of demand for such bricks, which did not cease even during the depression. In 1935, seven years after the appearance of the new bricks, of the 250 glass-melting furnaces operating in the U.S.A., only 20 remained without the use of the new material.

In observing the quality of production, the physicist establishes standards, develops methods for the necessary measurements, and supervises the carrying out of these measurements. Properties thanks to which the specific use of each kind of glass becomes possible, as well as properties that best control the uniformity of melts, are checked continuously.

For example, the electrical resistance of some glasses is very sensitive to changes in composition. The wires carrying current to an electric lamp enter the lamp through the stem—a short glass tube into which they are sealed. If the electrical resistance of the tube is not sufficiently high, the current is short-circuited through the glass instead of passing through the filament of the lamp. It is therefore necessary constantly to control the resistance of the glass going into the manufacture of the stem. The coefficient of expansion of this glass must also be controlled, since the stem is soldered to the wires and to the bulb. The coefficient of expansion must likewise be carefully controlled in the production of glass for the caps of railway lanterns, for chemical cups and other objects that must withstand changes of temperature without cracking.

Determination of the softening point of glass, which reduces to measuring viscosity, is a good control of the uniformity of successive glass melts and is important in itself for, for example, electric-lamp bulbs, which are produced in lamp factories by means of an automatic machine requiring that the glass soften in a given flame always to the same degree. A whole series of properties of glass determining its usefulness is subject to measurement. In the course of a year the number of such measurements reaches many thousands.

No matter how good it may be, glass can break, and it happens that the consumer intends to place responsibility for the breakage on the industrialist. Here again physics comes into its own and, on the basis of studying the surface of the fracture and interference in polarized light, draws its conclusions as to whether the article broke from impact, overheating or overcooling, or whether it had served out its time, etc. Often the physicist can propose changes in the design or in the method of using the glass that will eliminate further misunderstandings of this kind. Temperature measurements in the bulk of the melt first became widespread after, about 30 years ago, one physicist placed a platinum–platinum-rhodium thermocouple in a glass-melting furnace. Despite the stokers’ skill in determining temperature by eye, in fact the temperature of the furnace varied from the temperature at which the difficult-to-melt and disintegrating melts were melted to a temperature insufficient to melt the glass properly. The introduction of accurate temperature measurements eliminated this defect and at once gave a substantial increase in the productivity of the furnaces.

As a result of joint work by a physicist and a chemist on improving the composition of glass, a glass was created for railway lanterns that proved sufficiently resistant to rapid heating and cooling that the cap on a burning lantern does not burst during rain or snowfall. Such glass must have as small a coefficient of expansion as possible. It is still far from sufficient to calculate a composition suitable for these purposes. This glass must, along with a low coefficient of expansion, possess adequate fusibility, suitability for working, absence of a tendency to devitrification, resistance to the destructive action of water and other reagents, etc. Starting from a composition with known properties, the approach to the desired type of glass proceeds step by step, very gradually. At each point it is necessary to check what changes have occurred in the coefficient of expansion and in other properties of interest to us. Careful physical measurements are required in order to make sure that progress is proceeding in the desired direction.

The glass with a low coefficient of expansion developed in the USA especially for the needs of the railways found wide application in many other fields as well. The production of chemical, kitchen, and tableware would hardly have arisen without the aid of physics.

The physicist played an irreplaceable role in such a matter as the production of glass for lamp bulbs, which reproduces the properties of the flint glass formerly used for these purposes and, in addition, can be melted in a furnace. Thanks to this, the modern automatic production of bulbs became possible.

The recalculation of lenses for railway semaphores and the standardization of railway signal colors are another example of the physicist’s activity in creating a new improved product. Previously, each railway, independently of the others, chose a number of signal colors which, in this way, were never the same on any two roads. Such a situation was unsatisfactory both from the point of view of railway operation and from the point of view of glass production. The present standard (on the railways of the USA and some other countries) colored glasses were selected with the following requirements in mind: 1) their color must be the clearest; for example, yellow must be neither reddish nor greenish; 2) they must transmit as large a part as possible of the light emitted by the source. Furthermore, by reconstructing signal lenses on the basis of taking into account the actual shape and dimensions of the flame (instead of the former calculations for a theoretical point source), the physicist doubled the range of visibility of the signal.

The direction of the production of signal glasses is carried out by the physicist. A sample of each melt is subjected to photometric measurements, and if it does not fall within the prescribed limits with respect to

of color and transparency, the melt is not released for processing. Each piece of colored glass produced for signal lanterns is also checked photometrically.

The physicist has brought about a number of commercial applications of glasses with special electrical properties. On the basis of his own dielectric-strength measurements he introduced into use suspension and pin-type glass insulators for high-voltage lines and showed that they possess certain advantages. Together with his colleague, an electrical engineer, he designed an insulator of unsurpassed strength. For the transcontinental telephone line he proposed glass insulators with unusually low dielectric losses. Such glass has also found application in insulation for radio receivers and transmitters.

The physicist and the chemist, working in close cooperation, created glass filters that retain or, conversely, transmit individual regions of the spectrum from X-rays to infrared rays. The physicist’s role consisted in measuring the energy absorbed by, or transmitted through, the filter for rays of different wavelengths, while the chemist selected such batch compositions that the glasses obtained would not only be the required filters for radiant energy, but would also be normal in other respects.

Thus, for example, there are glasses for spectacles used by physicians for protection against X-rays and, on the other hand, glasses that specially transmit X-rays. One kind of glass transmits ultraviolet almost as well as quartz; another kind of glass (used for “mountain-sun” lamps) transmits only that region of the ultraviolet which has a therapeutic effect. The next kind of glass transmits ultraviolet while absorbing visible light, and is used in combination with a fluorescent material for stage effects. Glass of one kind, transparent and almost colorless, transmits practically all visible light and stops all heat rays; glass of another kind does not transmit visible rays but is sufficiently transparent to infrared rays. For air conditioning, glasses are used that weaken bright sunlight and absorb a large part of the heat rays, without cracking from heating because of their small coefficient of expansion. There are filters in combination with which the light of an incandescent lamp imitates direct sunlight or northern sky light, or cloudy sky. To all this the physicist has applied his hand.

The manufacture of frosted bulbs for electric lamps, which have now almost displaced transparent bulbs, as well as of globes and lampshades of milk glass, is based on physical measurements of the scattering and transmission of radiant energy.

Automobile rear-view mirrors, reflecting the light of the headlights of oncoming cars on the road and thus protecting the driver from their blinding effect, also appeared thanks to the physicist.

We shall also mention classic examples of cooperation between physicist and chemist in work with glass, although these physicists were rather university physicists than industrial ones. In the nineteenth century the chemist Harcourt with the physicist Stokes and later the chemist Schott with the physicist Abbe, on the basis of a scientific method of organizing production, developed a series of new formulas for optical glasses, as a result of which the modern microscope and other optical instruments with achromatic objectives could appear.

In the modern production of artificial fiber and fabric from glass, the physicist constantly controls the heat-insulating properties of the material. Thanks to the high electrical resistance of artificial fiber, the physicist was able to use it for the insulation of electric wire and cable. Measurements of the tensile strength of artificial fiber showed that such fiber, only a few ten-thousandths of an inch thick, has a strength exceeding that of steel

E. K. SULLIVAN

and reaching up to a million pounds per 1 dm². From this the physicist concluded that glass, by its inner nature, is a sufficiently strong material, and that the effort to obtain ever stronger glass products thus has real ground beneath it.

Thermal treatment of glass, guided by physics, is acquiring ever greater practical importance. Annealing, required for the removal of stresses that may cause spontaneous cracking, used to be carried out by rule of thumb and produced a rather high percentage of rejects until the physicist developed a special temperature regime of cooling.

As with steel, so with glass, thermal treatment affects a number of physical properties. The great viscosity of glass accounts for its tendency to slow down the internal rearrangements that usually accompany a change in temperature. When, say, water is cooled, the change in its density proceeds in parallel with the change in temperature. Glass, however, is so viscous a liquid that, if it can be called a liquid, the change in density and in other properties does not occur at once, but requires time. Thus, when glass is rapidly cooled to the temperature of solidification, the density and other properties may be caught in the process of change, and their numerical values, corresponding to a high temperature, may be as it were frozen into the cooled glass. With slow cooling, the internal rearrangements have time to follow the change in temperature, and thus a glass with normal properties is obtained. Glasses obtained by both of these methods, identical in their chemical composition, are equally free of elastic stresses, but their physical properties may differ by several tenths of a percent. This discovery sheds light on the difficulty of exact reproduction of glass. The thermal treatment of optical glass, for example, must be carried out with particular care in order to ensure that the required refractive index is obtained.

Thermal treatment of another kind is used to increase the strength of glass by 4–5 times. In Parisian taxicabs the windows are not the usual two panes of glass with a plastic interlayer between them, but a single sheet of thermally treated safety glass, which is 4–5 times stronger than ordinary glass of the same thickness and, if it does break, produces harmless fragments.

To obtain such strength, window glass is heated and then subjected to the cooling action of many air jets. The surface layer of the glass is thereby put into compression, while the inner layers are put into tension. The tensile strength of glass increases as the surface compression increases, since the force tending to tear the glass apart must first overcome this compression. Glass fibers, saucepans, and pots that can be heated directly over a gas flame or on an electric hot plate are also among the latest achievements in which the physicist has played an irreplaceable role.

The making of the disk for the mirror of the large telescope of the California Institute of Technology is almost entirely the work of the physicist’s hands. This disk, flat on one side and ribbed on the other, has a diameter of 70 feet and a thickness of 27 dm, and weighs 20 t—dimensions without precedent. It had to be made of borosilicate glass with a coefficient of expansion lower than any previously achieved. This required as much annealing as, say, glass for a microscope objective. There were no established methods for melting glass for these purposes or for giving it the necessary shape. At every step it was necessary to work out new methods and create special equipment. Although a defect in any one of the many details of planning or execution could have caused a long delay and considerable loss, the project was successfully brought to completion thanks to the skill and inventiveness of industrial physicists.

In addition to such purely practical questions, with the solution of which physicists are occupied, they are also working on purely scientific problems connected with glass; the solution of these problems, we are convinced, will be no less fruitful than the efforts applied today directly to production methods.

With this list of the activities and achievements of physics in the glass industry I have tried to show how great the role of physics is in establishing and maintaining quality standards, in overcoming production difficulties, and in seeking new ways of using glass.