896
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Vol. 18, s o . 9
Fifty Years of Glass-Making’ By Alexander Silverman DEPARTMENT OF CHBMSTRY, UNIVZRSITY OB PITTSBURGH, PITTSBUEGH, PA.
URING the half-century that has elapsed since the founding of the AMERICANCHEMICALSOCIETY many new industries have had their birth. Glassmaking, one of the oldest of the arts, probably had its origin in Egypt over 3500 years ago, and was the first industrial enterprise in the American Colonies, with a factory near Jamestown, in the Colony of Virginia, in 1607. Nevertheless, developments made in this field during the past fifty years eclipse everything that preceded. Even the application of art in glass manufacture and treatment has possibilities today which equal those of old, and excel, because larger scale production enables more of us to possess the product. Outstanding among the developments are modern window, plate, bottle, and hollow-ware manufacture, tube drawing, chemical and optical glass making. Inseparably coupled with many of these developments is the evolution of fuel, and furnace design and practice.
D
Window Glass
We are told that Theophilus, an Anglo-Saxon monk of the twelfth century, blew the first window-glass cylinder, cracking and flattening his cylindrical bubble, as is still done in the Belgian or hand process, except that the procedure was crude. Robinet, a Frenchman, used an air pump for blowing glass in 1824, and already in 1854 Loup, another Frenchman, resorted to the bait and compressed air for drawing window-glass cylinders. Real success did not come, however, until 1903 and after, when the Lubbers machine was invented. This, with many mechanical improvements which have been patented from time to time, is the machine which produces most of our cylinder glass today. I n principle it is the Loup idea, an iron bait dipping into molten glass, compressed air blowing the initial bubble, and air under lower pressure preventing collapse of the cylinder while i t is being drawn. A subsequent and novel modification by Frink should be mentioned. I n this the bait is a metal cylinder of substantially the same diameter as that of the glass cylinder, and air a t constant pressure is introduced through a raised opening in the middle of the crucible which contains the molten glass. Loup drew his cylinder right out of the pot in which the glass was melted. I n 1872 regenerative tank furnaces had already replaced some of the regenerative pot furnaces invented by Siemens and first used in Germany in 1861. These tank furnaces were a boon to window-glass manufacture, for it was only necessary to ladle the constantly forming glass into shallow crucibles, lower the bait, and blow and draw the cylinders. Then came the reversible double crucible, whose under side was exposed to flame for melting adhering glass away while the upper half contained molten glass which fed the glass cylinder. Of course, there were accessory inventions such as supports for lowering the huge cylinders, electrically heated bars and nichrome wires for cracking, rotating ovens for flattening, and continuous lehrs for annealing. The muffle lehr, where radiated heat insures greater uniformity than direct firing, is gradually gaining precedence, and when electricity is sufficiently cheap elec1 Received July 14, 1926. To be presented before the Division of Industrial and Engineering Chemistry at the 72nd Meeting of the American Chemical Society, Philadelphia, Pa., September 6 to 11, 1926.
trically heated lehrs will undoubtedly take the lead because of facility of kontrol. The slitting and flattening of cylinders seemed wasteful of time, and it was long the ambition of inventors to produce sheet glass directly. Two successful processes have resulted, the first invented by Emile Fourcault in Belgium, the second by our fellow-countryman, Irving W. Colburn. Fourcault secured his first patent in 1902; Colburn in 1903. They are entirely different in principle and operation. That the window-glass world has taken serious note of the importance of direct and continuous sheet production is evident from the number of patents issued weekly for processes or their improvement. The Fourcault process depends on glass rising through a slot in a fire-clay float, subsequently being carried upwards between rolls. The annealing is done in vertical lehrs as the sheet rises. I n the Colburn process the glass adheres to a straight metal float-rod or bar, which is carried upwards and then over rolls in a horizontal direction. Flames play on the sheet where it changes its direction from vertical to horizontal. A single Colburn machine is said to produce more glass than forty men formerly blew by the hand method. Wire glass is now manufactured by leading the wire between two sheets of glass. This has the advantage that, although it may crack, the glass does not shatter and fly. From what has been said, one is apt to think that all of the contributions to window-glass manufacture have been made by the engineer. Cylinder and sheet glass machines are worthless without physical and chemical control. Temperature control and uniform rate of draw are essential. The viscosity of the product must be constant. The glass must be clear and not have a tendency to devitrify. It must be free from seeds or gas bubbles. It shall not contain stones from the refractory lining of the furnace. It must be without cords or striae. The surface shall be bright when the sheet leaves the factory and must stay bright. These exacting requirements can only be met by careful physical and chemical control, and the fact that this is still seriously neglected in some plants is seen in the manufacturing losses and quality of product. A machine is no better than its product. If the machine is inflexible, the glass must be uniform in composition and worked under proper control. Plate Glass
When Lucas de Nehou first rolled plate glass at Saint Gobain in 1688, he li6tle dreamed of the vast production of a n “automobile age.” The French process was substantially that of the present day, consisting of pouring the glass on the casting table, rolling, annealing, grinding, and polishing. Plateglass manufacture began a profitable career a t Creighton, Pa., in 1881. Although the melting furnaces have improved and mechanical methods have largely replaced labor, it is only a few years since glass has been run from continuous tank furnaces, onto casting tables, through lehrs, and to the grinders and polishers, in one continuous process. This is virtually a third continuous sheet process, and differs from window-glass manufacture in that the glass is rolled instead of drawn. I n plate glass, as in window glass, wire may be introduced, but here a single plate is rolled, wire laid on; and then more
glass poured and rolled over the wire. I n tlie new continuous process the principle of running wire between plates may apply as in vindow-glass manufacture. Bullet-proof glass is another plate product. Here two or more plates of glass are cemented together under pressure by a transoarent nlastic material. This is of service in banks and armored cars. The manufacturer of rriate and window rlass is r e t t i w some
serve for tlie various sealed-in tubes, roda, and metal parts, without crystallizing or "devitrifying," &s the glass-maker would say. I n the manufacture of lamps enormous quantities of glass tubing and rod are employed. Machines now make both continuously. Then there is fused quartz, opaque to transparent, which enters into the manufacture of rnercury and other vapor lamps to provide ultra-violet light.. As early as 1839 Gaudin prepared fiber and shot of fused quartz, but it was not until recently that Elihu Thomson aiid his associates prepared a product of great transpareiicy and purity in a specially desigiked vacuim-pressure electric furnace. This quartz has promise for surgical lighting instruments, opt.ica1 devices, and window panes for admitting ultra-violet light. To be sure, one must not overlook the properties of glass substitutes of organic origin for the last-namd purpose, as the "urea-formaldehyde" sheet and products of older methods also transmit ultra-violet light. Liehting and Table Ware
The tendency in lighting has been to direct the light flux, through illumination studies and design, and to increase or
buibs for general <ing is the inside-frosted bulb. This required extensive research for its perfection. At first the bulh was subjected to a single etching and crushed like an eggshell under the vacuum. Later it was found that a second or even third wash with etching solution of a different composition would tone down the rough cracks and strengthen the glass. T h e s e b u l b s a r e rapidly displacing tlie older types, and possess the advautages of external cleanliness-and higher light transmission. X-ray bulbs, r e q u i r i n g glass of such a composition as to permit the generation and passage of the penetrating rays, are another product of the last h a l f - c e n t u r y . With these came glass shields to absorb the rays and protect the operator. The radio industry with its many spocial types of tubes has again required research. Glass had to be produced that would
decrease light transmission for escieucy or arb's sake, according t,o the ideas of the designer. When carbon lamps were employed, alabaster glasses were desirable, as opal glasses transmitted too yellow a light. With the advent of the tungsten lamp, especially the concentrated-fila~iient type, glare had to bo eliminat.ed without impairing transmission. Studics on high zinc oxide-alumina-borosilicate opals have given us a new type of glass for intense light sources. Signal lights have been the subject of much study and their colors are now largely standardized. Among the new coloring agents t h a t have played a leading role during t.he fifty years is selenium. It is not onlv used in signal lenses, but has entered society in the amber, orange, ruby, and pinkglasses which have been so popular for table ware. Selenium has also largely displaced manganese dioxide as a decolorizer for correcting the green color in glass which is due to iron compounds in the raw materials, while neodymium oxide accomplishes the same purpose io borosilicate glasses. T h e f l u o r e s c e n t greenish canary of uranium glass and golden yellow of cadmium sulfidehavea,lso beon popular. T h e r e a r e really few new coloring agents. Optical Glass
Glaes-Covered Streef in Loe hngeies. Calif., an Entire Block in Length
I n 1878 Ernst Abbe made an appeal for research on o p tical glass and in 1881 attracted the attention of Otto Schott. Eseeptfortrialswith boras and titanium oxide, only compounds of silicon, sodium, potassium, calcium, and lead had been employed in optical-glass manufacture prior t o t h a t time. I n S c h o t t ' s experiments the effects of compounds of more
898
INDUSTRIAL A N D ENGINEERISG CHEMISTRY
than twenty-five additional elements were studied. Already in 1891the Jena factory announced fifty glasses possessing new optical properties. The late war caught the United States napping in optical-glass manufacture, as in a number of other industries. Re i t said to the credit of the Geophysical Laboratory and our government bureaus that with the co6peration of manufacturers the situation was soon remedied. We are now able to meet our needs and have added the porcelain pot and large-scale manufacturing methods-for example, melting in tank furnaces and rolling optical glass in plates. Castings up to 250 centimeters in diameter are made in Europe and we are still limited to about 30 centimeters, but the rest will come. A number of special types of optical glass are now in demand. One of these contains cerium oxide, boric oxide, and ferrous or cobalt oxide, and is used to protect the eyes against ultra-violet and heat rays in high-temperature cutting and welding. Modifications give us the various tinted lenses sold to protect the eyes out of doors. Another type in sheet form is so constituted as to let the actinic rays through, thus approximating transparent fused quartz. Silicon-free glasses are in the experimental stage, and in one instance germanium oxide has been substituted for the silica. Chemical and Cooking Ware
Glass of low coefficient of expansion and high resistaiice to chemical corrosion was made in America in the years inimediately following the birth of our SOCIETY. Many of us recall Nonsol. Then came the Jena glass of Schott, and with the war Pyrex, the work of Taylor and others at Corning. More recently the laboratory glass, with slight modification, has been blown and pressed into cooking ware and baking dishes in a number of plants, and i t is now a staple article on the market. Not only are the dishes transparent, or nearly so, but the same may be said of panes in oven doors. Heat-resisting glass may be colored, and is used for lantern globes and signal lenses to eliminate the danger of breaking in inclement weather when there is a hot light source on one side of the glass and snow or rain on the other. Treatment and Decoration
The inside frosting of bulbs has been mentioned elsewhere in this article. Acid polishing is another interesting development applied in the manufacture of cut glass. Lead glass is ground, but instead of being polished with rouge is dipped into a mixture of concentrated hydrofluoric and sulfuric acids. A gelatinous scum forms, which on being brushed off leaves a bright surface. The air gun is used to apply decorative colors and metallized coats to glass. The silvering processes of Petitjean and Liebig have been improved and mirrors are now backed electrolytically with copper instead of paint. Furnaces and Accessories
I n the earlier days of glass-making raw materials were piled in any convenient place in the factory. Now they are stored in concrete bins, which insure greater uniformity as the contents are not exposed so freely. From these bins the materials go to modern scales, frequently automatic, then to mechanical mixers of the concrete type, and in window or bottle factories, by belt conveyor to the melting tanks. Furnaces are now largely of the regenerative or recuperative type, fired by natural gas, producer gas, or oil. Tank furnaces for window-glass manufacture hold several thousand tons of glass and may melt three to four hundred tons per day. A recent invention suggests the use of small melting compart-
Vol. 18, No. 9
ments which feed the finished glass into the larger tank, claiming facility in repairing the melting units, where most of the tank corrosion takes place. There is a further possibility that pot furnaces of the future will be of the small one- or two-pot type, because of the great variety of glasses now being manufactured and the consequent necessity for a variety of temperatures easily controlled, not only for the melting, but also for the subsequent working of the products. The old idea of using a standard batch and simply adding this, that, or the other coloring agent is fast disappearing, as many of the new coloring agents require special batch compositions. Extensive studies on the composition of refractories are under way, and a better understanding of the function of raw materials and nature of the physical and chemical changes taking place during their use is in prospect. The successful casting of pots and other refractory units has been accomplished, accompanied by economy of time in manufacture and drying. Pot-arch design has had considerable attention. The studied circulation of heat and the higher temperatures obtainable have effected economy through the more uniform heating of the pot with greater resistance to corrosion. Lehrs have already been mentioned. The muffle type is rapidly supplanting direct firing, and electrical annealing is only a question of cost of current. Combustion studies and pyrometric control have played a big role in the economics of glass manufacture. The plant that pays no attention to its fuel, to the nature of furnace gases, to temperatures as a check on furnace and lehr operation, is a back number. CalEulations
Last, but not least, in this list of topics discussing deyelopments of the past fifty years comes the subject of “calculations.” Data are accumulating, slowly but surely, which enable technical men to anticipate possibilities in glass manufacture, so as to minimize the cost of experimentation and supplant the rule-of-thumb methods by more reliable ones It is now possible to study the physical properties of a glass and tell how it shall be annealed. The strain-finder, employing polarized light, tells the manufacturer whether the glass has been annealed properly. The study of two glasses will determine whether they may be worked together, or how they must be modified to accomplish this. Trade secrets, which so strongly characterized the glass industry years ago, are things of the past. The chemist, who has also had experience in the factory, can analyze practically any glass that is submitted to him, and reproduce it in short order. The Trend of the Times
I n the early days of glass-making a manufacturer had a tendency to make anything the customer wanted, from a vase to a window pane. Today specialization prevails, one manufacturer making window glass, another plate glass, a third table ware, a fourth bottles, and so on. The introduction of automatic or semi-automatic machinery has made this possible, or perhaps necessary. I n the past small plants sprang up wherever there seemed to be a prospect of cheap fuel. Today a more careful study is made of future prospects. Fifty years ago the glass industry centered chiefly about Pittsburgh, central New Jersey, and New England. Today it is pretty well distributed, with a somewhat greater increase showing in California factories. Fifty years ago a midwestern plate-glass factory with over two thousand employees failed for lack of tariff protection. Today it is a function of a government commission to study-
INDUSTRIAL A N D ENGINEERING CHEMISTRY
September, 1926
the needs and afford protection. Not only are these studies important, but the economy effected by the industry itself, through consolidation, plays a part. One large company, which today has only six factories, a t one time operated fifty-three; yet with its six plants it produces about three times as much window glass, and of a much better quality. The grouping of factories in various lines, instead of having a great many small plants compete, has succeeded in stabilizing industry. Today there are three very large producers of window glass, four of bottles and containers, and two of plate glass. Table-ware manufacture is still scattered in many small units because automatic and semi-automatic machinery has not been applied so largely in this branch of the glass industry. Figures taken from United States census reports indicate the relative magnitude of the glass industry for 1879 and 1923. Values were not available for 1876 and 1926. 1879 Number of establishments Number of persons engaged Capital Salaries and wages Value of products ” For 1919, as 1923 figures were not
169 24,177 $18 805 000 $ 9’144’000 $21:155:000 available.
899
Science us. Art
Art naturally gives way to science during the early stages of mechanical development, and has done so in the glass industry, as in others. However, art, as a matter of mass production, thrives better in a mechanical age. So, while variety and individuality may be lacking in the products, what art there is can be produced cheaply and reaches a greater number of people. Real art, the “art for art’s sake,” after all appeals to but few producers. To these, American glass manufacture can point with real pride. Science has just begun to investigate the possibilities of glass-making. Her strides are becoming rapid and the prospects promising. Let us hope that her sister, Art, will not fall too far behind, and that Science will not forsake Art in the fascinating journeys into this field of endeavor. Acknowledgment
An article of this length, attempting to cover the progress of an important branch of manufacture during a period of fifty years, is a t best only an outline. Tendencies in the industry as a whole and in its leading branches, have, therefore, had brief treatment. For matter pertaining to business tendencies the writer is obligated to J. ill. Hammer.
1923 333 79,679 $215,680,486“ $105 417 374 b309:353:411
Water Supply and Sewerage during the Past Fifty Years’ By Robert Spurr Weston 14
BEACON
ST., BOSTON,Mass.
Water Supply
IFTY years ago there was a chemist, Pasteur, who had ended the purely mechanical period of water supply by connecting certain bacteria with certain diseases. His discoveries changed the art fundamentally. It was, of course, true that both good and bad waters existed then as they do now, but they were selected largely by appearance and cost, although it was also true, as it was in Galen’s time, that certain waters were rejected because of their corrosive action on lead. By 1876 Prof. William Ripley Nichols, the eminent head of the Chemical Department of the Massachusetts Institute of Technology and an authority on water, made several reports on the subject of plumbism, dealing with the action of water on lead and the protective action of autogenous coatings within lead service pipes. However, many public supplies were drawn from polluted ground and surface sources and, as a general rule, deep ground waters and surface waters from large lakes or upland reservoirs were the only safe sources. Kirkwood, the engineer of the St. Louis Water Works, had recently written about European water filters, and the first two of these in the United States had just been placed in operation at Poughkeepsie and Hudson, N. Y. I n Europe, the practice of filtration, begun in London in 1829, had spread through Northern Europe and in 1876 there was no German city supplied with unfiltered river water. The same was largely true in England, although storage rather than filtration was depended upon in many cases. Mechanical filters were unknown. Typhoid fever and cholera were devastating diseases and although vital statistics in the worst localities were not
F
1 Received
June 2, 1926.
kept until later, death rates were horrible to contemplate I n fact, the contagiousness of typhoid fever, following the work of William Budd (1873) and others, had only just been accepted, but not generally, and the typhoid bacillus had not been discovered. Thousands died of cholera yearly, and the protection afforded by filters against the cholera spirillum, not yet discovered by Koch, was not proved until the Hamburg epidemic in 1892. Water softening by the lime or soda-lime method was practiced quite extensively, but the methods were empirical and crude. The past art of water supply reflects the work of bacteriologists and analysts, but now is reflecting to an increasing degree the work of biological and physical chemists. Water purification methods now rest upon the excellent experimental foundations laid by the Massachusetts State Board of Health a t its Lawrence Station, and by such workers as Drown, Sedgwick, Mills, Hazen, Fuller, and Clark in this country, and Lindley, Houston, Thresh, Piefke, Kemma, and Halbertsma in Europe. The English filter as refined and standardized by Hazen and others is still employed, but the more economical and adaptable rapid filter standardized by Fuller and others, and using coagulants, is rapidly outstripping it in practice. I n 1876 only thirty-five thousand people in the United States were supplied with water purified by filtration. At present, in the United States alone more than as many million people are so supplied and from works having an output capacity of over five billion gallons daily. Many of these plants combine softening with purification. Fifty years ago the theory of filter action was simple. Professor Nichols mentions three effects--straining, “a kind of sedimentation,” and absorption. I n this day one
